Concepts
All Models
Auto-generated index of all models mentioned across the wiki.
All Models
Auto-generated index of models extracted from wiki entries. 1368 models with 2+ mentions.
ChatGPT
Mentioned in 1362 entries
- 1010-no-notes
- 12-months-apart
- 25x-increase-in-usage-of-our-agentic-products-codex-and-chat
- 3-love-how-it-understands-context-without-context
- 30-of-princeton-students-cheat-with-ai-but-peers-wont-snitch
- 3050-sessions-start-like-this
- 55-is-unbearable
- 56-high-is-answering-like-instant-it-seems-unable-to-actuall
- a-23-year-old-ucla-graduate-identified-as-andre-mai-went-vir
- a-chatgpt-generated-100-step-visual-guide-to-making-a-100-st
- a-comparative-analysis-of-institutional-and-course-generativ
- a-confusing-thing-about-the-proliferation-of-ai-modes-chatgp
- a-failure-of-chatgpt-work-claude-cowork-is-they-assume-that-
- a-funny-glitch-i-encountered-where-chatgbt-suddenly-said-a-s
- a-look-at-spiralism-a-quasi-spiritual-movement-that-grew-in-
Claude
Mentioned in 946 entries
- 1-the-huggingface-attack-was-a-felony-under-the-computer-fra
- 10-agents-hit-claude-at-160831-all-10-recovered-on-ollama-by
- 30k-likes-in-the-first-hour-that-is-crazy-design-is-unsolved
- 4-5-days-replacing-claude-w-qwen-38-next
- 45-still-came-in-030-under-claude-codes-spend-at-a-similar-s
- 6-claudes-27-products-1-voice-heres-the-rulebook
- 8-days-ago-while-jogging-i-asked-claude-to-solve-the-riemann
- a-bit-more-about-why-i-think-what-dawkins-wrote-is-misguided
- a-comprehensive-survey-of-agents-for-computer-use-foundation
- a-confusing-thing-about-the-proliferation-of-ai-modes-chatgp
- a-couple-of-days-ago-i-did-a-quick-explainer-on-claudes-new-
- a-critical-question-in-agent-design-is-how-do-we-build-agent
- a-failure-of-chatgpt-work-claude-cowork-is-they-assume-that-
- a-federal-appeals-court-upholds-dods-anthropic-blacklisting-
- a-few-weeks-ago-simonw-got-claude-to-port-liteparse-to-the-b
Codex
Mentioned in 385 entries
- 1password-codex-engineering-rollout
- 1password-extends-openai-collaboration-with-codex-mcp-server
- 20-ollama-vs-20-codex
- 25x-increase-in-usage-of-our-agentic-products-codex-and-chat
- 3blue1brown-httpsxcomverncrawfordstatus2042344703077859402s4
- 55-in-codex-is-so-good-for-non-coding-tasks-i-keep-assuming-
- a-confusing-thing-about-the-proliferation-of-ai-modes-chatgp
- a-fundamental-problem-with-extending-codexcoworkcode-to-all-
- a-little-bit-of-codex-in-my-life-a-little-bit-of-claude-code
- a-lot-of-papers-coming-out-are-still-focused-on-gpt-4-but-yo
- access-openai-models-and-codex-through-your-oracle-cloud-com
- actually-i-did-a-quick-test-i-like-the-chatgpt-workcodex-spl
- afaik-this-is-the-first-time-its-been-made-official-from-an-
- after-seeing-these-tweets-i-decided-to-try-it-out-on-my-own-
- agent-quest-now-tells-you-when-claude-code-or-codex-needs-yo
Stable Diffusion
Mentioned in 366 entries
- 128gb-mac-studio---help
- 25d-fantasy-style-lora-for-anima-trained-in-1-hour
- 2k-anima-image
- 3-years-of-training-with-ai-tools-finally-put-to-use
- 300-prompts-across-photography-cinematography-lighting-and-c
- 5700-xt
- 90-second-reels-to-20-mins-docos
- a-comprehensive-dataset-for-human-vs-ai-generated-image-dete
- a-glimpse-into-the-fleet-deep-space-patrol-4k-sci-fi-oc
- a-road-movie-through-stable-diffusion-valley
- abandon-place---01
- acestep-15---lora-trained-on-first-two-albums-of-modern-talk
- acestepcpp-can-now-outpaint
- add-any-kind-of-audiovoicessfx-ambience-to-existing-video
- add-audio-from-text-prompt-to-existing-video
Llama
Mentioned in 315 entries
- 3-days-benchmarking-most-llamacpp-flags-on-my-weird-40gb-vra
- 50-tg-increase-with-offloading-hot-experts-to-vram
- 80-toksec-and-128k-context-on-12gb-vram-with-qwen36-35b-a3b-
- a-124b-emitted-15128-tokens-in-a-single-response-on-one-dgx-
- a-comprehensive-survey-of-agents-for-computer-use-foundation
- adding-logit-penalty-for-wait-maybe-and-perhaps-to-qwen-mode
- after-pushing-1m-tokens-through-qwen-38-27b-here-is-my-optim
- alibabas-open-weight-ai-models-accumulated-3-billion-global-
- am-i-just-hallucinating
- android-app-ollama-talk-now-on-playstore
- astra-is-awesome-but-if-youre-using-it-specifically-to-parse
- automegakernel-a-statically-checked-agent-harness-for-self-r
- b10083
- b10087
- b10106
Gemini
Mentioned in 285 entries
- 3-new-ways-ads-advisor-is-making-google-ads-safer-and-faster
- 8-gemini-tips-for-organizing-your-space-and-life
- a-comprehensive-survey-of-agents-for-computer-use-foundation
- a-few-weeks-ago-we-asked-our-community-to-use-googleaistudio
- a-little-bit-of-codex-in-my-life-a-little-bit-of-claude-code
- a-primer-on-computational-semantics-for-artificial-intellige
- a-reliability-assessment-of-lalm-audio-judges-for-full-duple
- a-survey-on-foundation-models-for-personalized-federated-int
- accelerate-postgresql-migrations-using-gemini-in-database-mi
- accelerating-scientific-research-with-gemini-in-the-real-wor
- adobe-canva-and-capcut-announce-gemini-integrations-to-let-u
- agents-dont-paginate-first-chunk-selection-for-llm-tool-resp
- ai-models-are-terrible-at-betting-on-soccerespecially-xai-gr
- alphaevolve-how-our-gemini-powered-coding-agent-is-scaling-i
- an-opinionated-guide-to-which-ai-to-use-to-do-stuff
CLIP
Mentioned in 244 entries
- a-classifier-agnostic-zero-shot-adversarial-attack-detection
- a-controlled-study-of-clip-based-body-scene-fusion-for-emoti
- a-novel-framework-for-open-vocabulary-multi-object-recogniti
- a-paragraph-is-worth-a-thousand-captions-rethinking-text-sup
- a3b2-adaptive-asymmetric-adapter-for-alleviating-branch-bias
- adversarial-attacks-already-tell-the-answer-directional-bias
- agc-adaptive-geodesic-correction-for-adversarial-robustness-
- aligning-forest-and-trees-in-images-long-captions-for-visual
- anchorscore-a-clip-based-diagnostic-of-mllm-annotation-diffi
- anomalyagent-training-free-agentic-models-for-zero-few-shot-
- area-attribute-extraction-and-aggregation-for-clip-based-cla
- aspectclip-optimizing-clip-representation-space-via-aspect-g
- attributes-should-come-from-images-not-class-names-distribut
- backdoor-as-probe-test-time-adversarial-defense-for-clip
- beyond-false-stability-high-noise-drift-gating-for-test-time
Grok
Mentioned in 199 entries
- 40-grok-build-agents-tearing-through-c-code-in-parallel-all-
- a-93-year-old-lady-filmed-her-tesla-fsd-taking-her-to-church
- a-survey-on-foundation-models-for-personalized-federated-int
- agi-ftw-ft
- ai-models-are-terrible-at-betting-on-soccerespecially-xai-gr
- and-major-improvements-coming-to-imagevideo-generation-accur
- and-now-grok-is-telling-the-cretins-to-stfu-this-is-an-enorm
- appmagic-grok-downloads-fell-to-83m-in-april-from-a-high-of-
- as-grok-flounders-spacex-bets-future-on-beating-big-tech-at-
- as-grok-gets-better-it-has-less-and-less-patience-for-the-cr
- asked-grok-one-simple-question-what-is-your-prime-directive-
- at-his-first-cursor-all-hands-musk-said-grok-needs-to-catch-
- banger-grok
- based-grok-httpsxcomigrokshare32212cc499ae467ebb1f8db2b77d31
- been-using-grok-build-these-past-few-days-and-the-thing-that
GPT-5.5
Mentioned in 164 entries
- 1-we-believe-in-iterative-deployment-although-gpt-55-is-alre
- 55-has-solved-it
- a-27b-agent-just-beat-claude-opus-48-and-gpt-55-on-held-out-
- a-pelican-for-gpt-55-via-the-semi-official-codex-backdoor-ap
- after-using-glm-52-for-a-day-im-surprised-by-how-often-it-fe
- an-analysis-of-gpt-55-gemini-31-pro-grok-43-gabs-arya-and-ot
- banger-paper-on-self-evolving-ontologies-for-agents-you-just
- but-but-i-thought-it-was-magic
- chatgpt-55-is-here
- claude-opus-48-benchmark-numbers-vs-gpt-55-are-kinda-concern
- cognition-releases-swe-17-trained-from-kimi-k27-and-availabl
- cost-per-successful-task-benchmarking-kimi-k3-gpt-55-and-8-m
- cursor-composer-25s-is-318x-cheaper-than-opus-47-in-claude-c
- cursors-new-composer-25-takes-third-on-the-artificial-analys
- databricks-brings-gpt-55-to-enterprise-agent-workflows
Mamba
Mentioned in 154 entries
- a-hybrid-mamba-for-audio-visual-navigation
- a-mamba-based-multimodal-network-for-multiscale-blast-induce
- a-simple-state-space-model-excels-at-multivariate-time-serie
- a-unified-mamba--moe-surrogate-for-closed-loop-simulation-an
- adamamba-adaptive-frequency-gated-mamba-for-long-term-time-s
- advancing-intelligent-sequence-modeling-evolution-trade-offs
- an-approach-with-visual-and-tabular-mamba-to-multimodal-medi
- an-exact-instrument-for-state-usage-in-selective-state-space
- an-exploration-of-mamba-for-speech-self-supervised-models
- an-ultra-widefield-swept-source-octa-dataset-and-a-polar-gat
- asymmetric-virtual-memory-paging-for-hybrid-mamba-transforme
- attention-mamba-a-mamba-enhanced-multi-scale-parallel-infere
- attention-to-mamba-a-recipe-for-cross-architecture-distillat
- avis-mamba-adaptive-visual-steering-of-audio-state-space-dyn
- b11060
Qwen
Mentioned in 151 entries
- 1-day-in-and-i-feel-okay-saying-muse-glimmer-30b-finally-bea
- 12gb-vram-gang-whats-our-plan
- 260816157-freetoken-efficient-edge-native-moe-serving-with-b
- a-great-example-of-practical-ai-solving-real-world-problems-
- ace-step-15-xl-comfyui-automation-workflow-without-lama-for-
- adding-logit-penalty-for-wait-maybe-and-perhaps-to-qwen-mode
- agenticqwen-training-small-agentic-language-models-with-dual
- alibaba-says-its-qwen-ai-models-will-be-integrated-into-byd-
- alibabas-open-weight-ai-models-accumulated-3-billion-global-
- alibabas-qwen-models-have-made-it-an-ai-powerhouse-but-the-c
- alternative-history-made-with-a-qwen-image-setup
- an-effective-rank-audit-of-alignment-induced-activation-shif
- anthropic-releasing-open-source-demos-built-on-top-of-qwen-w
- are-gui-agents-focused-enough-automated-distraction-via-sema
- atelier-prompt-studio-20-beta-a-free-local-prompt-crafter-fo
Hermes
Mentioned in 148 entries
- 384-people-in-the-hermes-agent-jam-session-in-our-discord-ha
- 3blue1brown-httpsxcomverncrawfordstatus2042344703077859402s4
- 4x-3090-96gb-vram-what-model-to-drive-hermes
- a-living-map-of-everything-your-hermes-agent-has-learned-eve
- actual-is-a-local-inference-stack-optimized-to-let-you-utili
- added-a-new-built-in-skill-inspired-by-claude-codes-simplify
- announcing-hermes-agent-community-meetup-in-seoul-south-kore
- anyone-here-using-ai-agent-orchestration-software-to-control
- as-always-hermes-update
- browser-extension-for-hermes-agent-nousresearch-hermes-lives
- call-your-hermes-agent
- coming-soon-worldsim-in-your-hermes-agent
- creative-direct-with-hermes
- deepseek-harnness---why-is-feels-better
- deepseek-v4-pro-is-now-on-ollamas-cloud-try-it-with-claude-c
GPT
Mentioned in 135 entries
- a-geometric-relation-of-the-error-introduced-by-sampling-a-l
- a-language-guided-bayesian-optimization-for-efficient-lora-h
- a-series-of-images-gpt-generated-of-ryuk
- a-third-grade-girl-used-gpt-to-generate-an-image-and-then-cr
- added-a-deepseek-sparse-attention-dsa-from-scratch-implement
- after-seeing-the-new-mortal-kombat-movie-i-asked-gpt-to-gen-
- ai-unpredictability-today-gpt-did-not-obey-a-permanent-instr
- aight-gpt-chill-tho
- all-benchmarks-are-flawed-but-gpqa-has-been-fairly-consisten
- all-posts-claiming-to-show-gpt-doing-something-wonky-should-
- almost-all-llm-thought-is-derivative-below-just-sounds-like-
- any-order-gpt-as-masked-diffusion-model-decoupling-formulati
- april-2026-newsletter
- bridgewater-just-published-numbers-that-should-make-every-fr
- browser-extension-for-hermes-agent-nousresearch-hermes-lives
Vision Transformers
Mentioned in 135 entries
- a-mechanistic-analysis-of-adversarial-fine-tuning-of-vision-
- a-smaller-transformer-in-your-transformer
- a-unified-framework-for-vision-transformers-equivariant-to-d
- accelerating-vision-transformers-with-adaptive-patch-sizes
- active-spatial-guidance-eliminating-injected-positional-mech
- adamerge-salience-aware-adaptive-token-merging-for-training-
- adaptive-hebbian-memory-routing-in-vision-transformers-for-f
- adversarial-decoys-misdirecting-attention-based-defenses-in-
- agrikd-cross-architecture-knowledge-distillation-for-efficie
- ai-generated-image-recognition-via-fusion-of-cnns-and-vision
- amortized-precision-quantization-for-early-exit-vision-trans
- an-open-source-two-stage-computer-vision-pipeline-for-fine-g
- asap-attention-sink-anchored-pruning
- attend-what-matters-leveraging-vision-foundational-models-fo
- attn-fiqa-interpretable-attention-based-face-image-quality-a
Qwen 3.8 27B
Mentioned in 113 entries
- 12gb-vram-gang-whats-our-plan
- 2x-tesla-p100s-q6k-quant-qwen-38-27b-60tps-v30
- 8-uncensored-qwen-38-27b-variants-one-base-167-gpu-hours---a
- 95-tps-through-100k-generated-for-qwen38-27b-262k-ctx-on-a-s
- after-pushing-1m-tokens-through-qwen-38-27b-here-is-my-optim
- am-i-doing-something-wrong-qwen-38-27b-seems-useless-for-age
- anyone-else-doing-egpus-oculink
- anyone-managed-to-get-qwen-38-27b-running-smoothly-on-vllm-c
- arex-2-vs-swift-vs-qwen38-27b
- artificial-analysis-intelligence-a-meaningless-benchmark
- baaiarex-2---27b---agent-model-based-on-qwen38-27b
- bro-wtf-qwen-lab-cooked-with-qwen-38-27b-its-so-fucking-good
- byteshape-qwen-38-27b-to-kl-diverge-or-not-to-kl-diverge-par
- can-a-relatively-small-language-model-qwen-38-27b-understand
- closed-ai-has-been-real-quiet-since-qwen-38-27b-dropped
Gemma 4
Mentioned in 111 entries
- 1-to-this-i-was-recently-on-a-cross-continental-flight-witho
- 128-gb-of-memory-is-nice-but-you-can-get-started-with-local-
- a-completely-local-agent-that-lives-right-inside-your-browse
- a-playable-digital-piano-built-with-gemma-4-e2b-2b-httpsxcom
- agent-factory-recap-how-gemma-4-taught-itself-physics
- agentic-kernel-optimization-is-the-future-of-on-device-infer
- amy-info-about-the-audio-capabilities-of-the-gemma4-models-i
- android-studios-native-gemma-4-runs-on-llamacpp
- april-was-a-pretty-strong-month-for-llm-releases---gemma-4--
- are--mlx-variants-equivalent-to--nvfp4-for-gemma4-and-qwen36
- are-1b-llms-going-away-in-2026
- b11065
- b11132
- b11238
- b8744
Kimi K3
Mentioned in 110 entries
- a-joint-preliminary-evaluation-by-the-uks-aisi-and-the-us-ca
- a-profile-of-yang-zhilin-founder-of-moonshot-ai-which-faced-
- alibaba-says-its-24t-parameter-qwen38-max-tops-moonshots-kim
- all-frontier-ai-models-including-gpt-56-sol-fable-5-and-kimi
- amazing-deep-dive-on-k3
- another-day-and-another-full-frontier-model-running-on-your-
- artificial-analysis-deepseeks-v4-flash-costs-0141m-input-and
- as-shocking-as-the-kimi-k3-release-massive-performance-gain-
- be-sure-to-tune-in-this-wednesday
- breaking-kimi-k3-by-kimimoonshot-is-1st-overall-on-3d-design
- bro-i-need-you-to-focus
- closed-ai-has-been-real-quiet-since-qwen-38-27b-dropped
- code-arena-now-measures-fullstack-capabilities-view-overall-
- cost-per-successful-task-benchmarking-kimi-k3-gpt-55-and-8-m
- deepseek-launches-v4-pro-its-most-advanced-model-that-rivals
Claude Code
Mentioned in 98 entries
- a-deep-dive-into-how-anthropics-claude-code-and-peter-steinb
- agent-quest-now-tells-you-when-claude-code-or-codex-needs-yo
- agentic-coding-in-the-wild-characterizing-github-copilot-tra
- agents-dont-paginate-first-chunk-selection-for-llm-tool-resp
- an-update-on-recent-claude-code-quality-reports
- apple-may-be-planning-to-role-out-its-updated-siri-based-on-
- as-the-progenitor-of-the-agent-lab-thesis-which-got-the-eval
- chatgpt-work-claude-cowork-chatgpt-codex-claude-code-i-kinda
- claude-code-is-not-agi-but-it-is-the-single-biggest-advance-
- claude-code-just-stopped-a-ddos-attack-on-bridgemind-in-unde
- claude-code-still-not-agi-but-biggest-advance-since-gpt-4-is
- claude-desktop-httpsxcomdoteystatus2055777343222808744-opena
- codegrep-an-rl-trained-retrieval-agent-for-llm-coding-agents
- consistent-with-what-i-found-with-qwen36-a-while-back-claude
- context-matters-repository-aware-security-analysis-of-the-ag
Transformer
Mentioned in 96 entries
- a-hormone-inspired-emotion-layer-for-transformer-language-mo
- a-multimodal-label-forecasting-method-for-aperiodic-visuo-mo
- a-quantum-roadmap-for-softmax-attention-exact-born-rule-anal
- a-simple-transformer-pipeline-for-full-key-side-channel-atta
- a-systematic-benchmark-of-machine-transliteration-models-for
- a-systematic-investigation-of-molecular-encoding-methods-for
- a-unified-geometric-space-for-topological-alignment-between-
- ame-ts-anchored-mixture-of-experts-for-time-series-forecasti
- asymmetric-virtual-memory-paging-for-hybrid-mamba-transforme
- at-the-edge-of-understanding-sparse-autoencoders-trace-the-l
- athena-knowledge-guided-agentic-neural-architecture-search-f
- attention-guided-dual-stream-learning-for-group-engagement-r
- attention-mamba-a-mamba-enhanced-multi-scale-parallel-infere
- attention-to-mamba-a-recipe-for-cross-architecture-distillat
- attentioncap-transformer-based-capacitance-matrix-learning-t
LTX 2.3
Mentioned in 93 entries
- a-fully-character-driven-fantasy-story-made-entirely-with-lt
- a-little-ltx-23-test-shake-them-gp-double-ts
- a-look-at-the-video-outpainting-workflow-in-comfyui-using-lt
- a-rushed-wedding-gift-for-my-brothers-chicago-blues-track-hy
- all-mt-ltx-23-outputs-look-like-this-no-matter-what-models-i
- anyone-figured-out-ltx-23-lora-training-yet-training-with-de
- apparently-this-clip-is-too-spicy-so-lets-try-it-this-way-ex
- are-there-face-swaps-or-head-swaps-inside-ltx-23-yet
- beginner-prompting-guide-for-ltx-23-tips-and-tricks
- best-ltx-23-model-format-for-rtx-3090ti-24gb-in-comfyui-movi
- best-ltx-23-sampler-scheduler-combo
- buy-rtx-5090-or-rent-h100-for-ltx-23
- can-ltx-23-do-romance-with-kissing-and-clothed-makeout
- character-workflow-chroma1-hd-flux2-dev-wan-22-ltx-23
- children-of-your-mistakes-ltx-23-chromahd-zimag-indextts2
Qwen3.8-27B
Mentioned in 93 entries
- 1-open-weight-model-on-harveys-legal-agent-benchmark-strong-
- 27b-on-17gb-ram-are-you-ready-to-create-something-incredible
- 4-days-to-the-topthank-you-to-every-builder-who-pushed-qwen3
- a-gift-for-developers-qwen38-27b-running-locally-on-amd-from
- a-hunch-qwen38-27bs-general-knowledge-got-pruned-good-if-tru
- a-local-27b-model-scoring-frontier-performance-huge-thanks-t
- a-preliminary-qwen38-27b-model-card-is-live
- airllm---recent-updates---with-qwen38-27b-kimi-k3-too
- alibaba-releases-weights-for-qwen38-models-under-apache-20-l
- alibaba-says-its-24t-parameter-qwen38-max-tops-moonshots-kim
- alibaba-says-its-new-open-source-multimodal-model-qwen38-27b
- bartowskiqwen38-27b-gguf-hugging-face---updated-per-tensor-l
- benchmark-results-what-is-the-best-and-fastest-engine-to-run
- cut-qwen38-27b-reasoning-tokens-by-40----38-thinkingcap-benc
- does-an-ollama-ready-build-keep-this-27bs-mtp-and-vision-sta
GPT-6 Astra
Mentioned in 92 entries
- ai-for-science-with-gpt-6-astra-thermal-design-and-electroth
- alongside-the-launch-of-gpt-6-astra-were-committing-1-billio
- also-this-is-my-favorite-openai-video-so-far-it-makes-me-exc
- an-unexpected-robot-policy-early-evaluations-of-gpt-6-astra-
- and-as-a-scientist-its-disappointing-that-we-dont-yet-know-m
- another-neat-one-i-had-gpt-6-astra-build-a-model-of-a-famous
- anyone-who-cared-about-the-science-of-ai-would-find-this-int
- artificial-analysis-coding-agent-index-gpt-6-astra-scored-67
- artificial-analysis-says-gemini-4-argon-high-matches-gpt-6-a
- astra-for-law-frontier-intelligence-built-for-your-practice-
- blend-url-viewer
- built-this-yesterday-with-qwen38-flash-next-nvfp4-262k-conte
- by-declaring-that-gpt-6-astra-has-ushered-in-the-agi-era-ope
- claude-opus-50-identified-the-ip-sharing-route-and-declined-
- correction-the-switching-costs-werent-enough-to-stop-my-desi
Claude Fable 5
Mentioned in 92 entries
- a-til-on-using-httpagentsviewio-to-calculate-token-spending-
- after-spooking-trump-into-safety-testing-anthropic-ai-models
- ainews-anthropic-claude-fable-5-mythos-but-safe-with-controv
- an-underrated-part-of-glm-53-is-its-low-hallucination-rate-c
- and-the-video-for-reference-httpsxcomclaudedevsstatus2064399
- anthropic-backtracks-on-a-policy-limiting-claude-fable-5s-ab
- anthropic-releases-its-first-mythos-class-model-claude-fable
- anthropic-says-claude-fable-5-uses-conservative-safety-class
- anthropic-says-the-department-of-commerce-has-lifted-export-
- anthropic-sets-ai-performance-records-with-new-mythos-5-fabl
- anthropic-updates-claude-fable-5s-biology-safeguards-to-redu
- anthropic8217s-long-sidelined-fable-5-is-greenlit-to-return
- anthropics-new-model-fable-will-silently-handicap-work-on-ll
- appreciate-it-now-lets-understand-the-world-through-the-eyes
- breaking-anthropic-just-dropped-claude-fable-5this-is-mythos
GLM-5.2
Mentioned in 87 entries
- a-year-ago-this-would-have-been-an-obvious-closed-model-task
- after-using-glm-52-for-a-day-im-surprised-by-how-often-it-fe
- as-token-usage-explodes-model-choice-becomes-product-strateg
- b10174
- bit-equivalent-on-policy-rl-for-glm-52-has-been-achieved-int
- build-your-own-claude-code-with-deep-agents-good-article-by-
- but-heres-the-punchline-normalized-to-90-cache-hit-rate-glm-
- cant-merge-claude-max-subscription-opus-48-with-zai-glm-521m
- chinas-zai-claims-it-can-match-mythos-on-cybersecurity
- cost-per-iteration-is-the-unlock-glm-52-on-together-ai-can-g
- do-you-know-what-you-pay-for-in-agentic-workloads-cached-tok
- early-sparks-of-rsi-ali-taha-waterloointern-explains-how-htt
- everyones-trying-to-find-where-to-test-glm-52-you-can-try-it
- glm-52-and-ikllamaccp
- glm-52-has-been-stuck-at-no2-on-hugging-face-trending-for-th
Grok 4.5
Mentioned in 86 entries
- 1-day-in-and-i-feel-okay-saying-muse-glimmer-30b-finally-bea
- added-grok-45-to-the-harbor-town-playable-gallery-here-is-it
- ainews-spacexai-launches-grok-45-first-opus-class-model-post
- and-grok-46-comes-out-in-a-week
- announcing-grok-45-our-first-model-trained-specifically-for-
- artificial-analysis-assessment
- as-usual-legacy-media-is-misrepresenting-the-situation-i-jus
- based-on-strong-positive-feedback-from-customers-in-our-beta
- breaking-grok-45-has-climbed-to-2-on-the-frontierswe-benchma
- breaking-grok-45-high-ranks-1-on-the-highwalk-benchmark-whic
- breaking-grok-45-just-claimed-the-top-spot-on-the-new-highwa
- breaking-grok-45-ranked-1-on-laurenbench-with-a-score-of-569
- built-from-scratch-by-grok-45-grok-build-in-ue58-a-cyberpunk
- built-with-two-prompts-and-goal
- congrats-to-spacexai-on-grok-45-trained-on-nvidia-gb300-nvl7
3D Gaussian Splatting
Mentioned in 85 entries
- 2xplat-decoupling-geometry-and-appearance-modeling-for-feed-
- 3d-gaussian-splatting-for-efficient-retrospective-dynamic-sc
- 3d-skew-gaussian-splatting-with-any-camera-trajectory-visual
- 3d-skew-normal-splatting
- 3deditsafe-defending-3d-editing-pipelines-from-unsafe-genera
- 3dgs-sc-a-controlled-static-screen-content-benchmark-for-3d-
- active-view-selection-with-perturbed-gaussian-ensemble-for-t
- artifactworld-scaling-3d-gaussian-splatting-artifact-restora
- artifixer-enhancing-and-extending-3d-reconstruction-with-aut
- asyncevgs-asynchronous-event-assisted-gaussian-splatting-for
- bake-it-till-you-make-it-ultrafast-spatial-texture-atlas-spl
- beyond-a-single-light-a-large-scale-aerial-dataset-for-urban
- bi-flowgs-bridging-generative-view-completion-and-gaussian-g
- bringing-a-personal-point-of-view-evaluating-dynamic-3d-gaus
- casdeblurgs-cascaded-2d-to-3d-multi-view-consistency-for-3d-
GPT-5.6 Sol
Mentioned in 84 entries
- 56-sol-much-better-in-chat-now-and-unlimited-text-chat-for-f
- a-few-days-ago-anthropic-shared-this-brilliant-prompt-i-was-
- accessing-openai-models-on-amazon-bedrock-from-australia-wit
- advancing-the-price-performance-frontier-with-gpt56
- after-deployment-we-applied-gpt-56-sol-to-advance-the-fronti
- ainews-openai-gpt-56-sol-terra-luna-restricted-to-trusted-pa
- all-frontier-ai-models-including-gpt-56-sol-fable-5-and-kimi
- an-internal-version-of-our-next-major-model-produced-10-new-
- and-the-balance-returned-to-earth
- artificial-analysis-deepseeks-v4-flash-costs-0141m-input-and
- as-we-continue-to-push-the-frontier-of-capabilities-while-im
- check-out-how-ai-helps-scientists-with-quantum-computing-res
- datasette-10a39-and-0654-security-releases---httpsdatasettei
- deepseek-v4-pro-0813-vs-gpt-56-sol-on-deepswe-cost-coding-an
- deepseek-v41-flash-has-arrived-on-together-ai-it-beats-gpt-5
Vision Transformer
Mentioned in 80 entries
- a-graph-augmented-knowledge-distillation-based-dual-stream-v
- a-hybrid-framework-of-vision-transformer-and-gated-recurrent
- a-mixture-of-experts-vision-transformer-for-high-fidelity-su
- a-survey-of-adversarial-efficiency-degradation-for-vision-tr
- a-universal-vision-transformer-for-fast-calorimeter-simulati
- a-vision-transformer-for-ecg-based-detection-of-left-ventric
- aa-vit-anatomically-aware-vision-transformer-with-structural
- accelerating-vision-foundation-models-with-drop-in-depthwise
- advancing-vision-transformer-with-enhanced-spatial-priors
- aegis-a-multi-task-joint-embedding-predictive-architecture-f
- an-interpretable-vision-transformer-framework-for-automated-
- artificial-intelligence-application-in-lymphoma-diagnosis-wi
- at-vit-area-targeted-multi-view-vision-transformer-with-cros
- attention-transfer-is-not-universally-effective-for-vision-t
- braided-vision-transformer-for-stroke-detection-in-multi-vie
Fable 5
Mentioned in 80 entries
- a-red-team-study-of-anthropic-fable-5-opus-48-models
- agentic-kernel-optimization-is-the-future-of-on-device-infer
- ainews-sonnet-5-today-and-fable-5-tomorrow
- all-frontier-ai-models-including-gpt-56-sol-fable-5-and-kimi
- am-i-supposed-to-interpret-that-its-better-than-fable-5-from
- anthropic-launches-claude-opus-5-which-it-says-comes-close-t
- anthropic-releases-opus-5-with-close-to-fable-5s-capabilitie
- anthropic-says-fable-5-has-invisible-safeguards-that-use-pro
- anthropic-says-fable-5-is-available-on-pro-max-team-and-seat
- anthropic-says-fable-5-will-be-available-via-usage-credits-f
- anthropic-says-fable-51-will-cost-an-estimated-25-less-than-
- anthropic-says-red-team-tests-of-fable-5-found-no-universal-
- anthropic-says-some-routine-tasks-like-coding-and-debugging-
- anthropic-says-the-us-government-is-allowing-mythos-5-to-be-
- anthropic-says-these-topics-are-too-dangerous-to-let-its-fab
ViT
Mentioned in 77 entries
- a-scalar-per-patch-from-pre-trained-vits-enables-fast-moving
- a-survey-of-adversarial-efficiency-degradation-for-vision-tr
- a2-smaller-self-supervised-vits-localize-better-than-larger-
- accelerating-vision-transformers-with-adaptive-patch-sizes
- advancing-vision-transformer-with-enhanced-spatial-priors
- adversarial-decoys-misdirecting-attention-based-defenses-in-
- an-automated-multimodal-glaucoma-detection-framework-using-v
- attention-transfer-is-not-universally-effective-for-vision-t
- audiofuse-unified-spectral-temporal-learning-via-a-hybrid-vi
- beyond-vit-tokens-masked-diffusion-pretrained-convolutional-
- bridging-modalities-and-tasks-a-unified-hierarchical-vit-for
- bridging-topology-and-deep-representation-learning-a-tda-vit
- bvit-investigating-single-block-recurrence-in-vision-transfo
- clear-moe-shared-basis-expert-extraction-from-frozen-vision-
- co-annotator-expert-distilled-vit-and-vlm-for-visual-and-doc
Diffusion Transformers
Mentioned in 76 entries
- accelaes-accelerating-diffusion-transformers-for-training-fr
- adacorrection-adaptive-offset-cache-correction-for-accurate-
- an-evaluation-framework-for-generating-multi-view-images-of-
- appearance-pointers----multimodal-region-control-of-diffusio
- attention-sparsity-is-input-stable-training-free-sparse-atte
- avits-adaptive-spatiotemporal-token-selection-for-efficient-
- awakening-diffusion-transformers-eliciting-stronger-generati
- bag-budget-aware-gating-for-diffusion-caching
- beyond-fixed-formulas-data-driven-linear-predictor-for-effic
- beyond-future-prediction-denoising-as-generative-adaptation-
- beyond-point-wise-matching-structural-representation-alignme
- beyond-text-conditioning-a-systematic-study-of-mllm-dit-fusi
- blocked-gibbs-meets-diffusion-transformers-unsupervised-lear
- brace-taming-sharp-irregularities-via-barycentric-rational-f
- calibri-enhancing-diffusion-transformers-via-parameter-effic
GLM 5.2
Mentioned in 75 entries
- 52-could-be-better-with-more-rl
- a-tangible-comparison-of-glm-performance-stacked-against-opu
- also-hearing-from-other-customers-that-glm-52-approaches-gpt
- and-grok-46-comes-out-in-a-week
- announcing-glm-arena-a-series-of-tests-infographics-svgs-sit
- as-shocking-as-the-kimi-k3-release-massive-performance-gain-
- breaking-grok-45-ranked-1-on-laurenbench-with-a-score-of-569
- cancelling-my-subscription-also-it-was-great
- claude-opus---glm-52-in-under-10-seconds-dcode-isnt-just-pro
- closed-ai-has-been-real-quiet-since-qwen-38-27b-dropped
- codex-is-overoptimised-for-large-models-it-ranks-2nd-out-of-
- dcode-does-a-bunch-of-work-to-standardize-message-format-mak
- desert-island-survival-list-solar-panel-battery-256-gb-mac-s
- european-inference-providers-for-glm-52-deepseek-v4-flash
- for-ollama-cloud-20-plan-what-models-are-you-guys-using
Kimi K2.6
Mentioned in 74 entries
- ai-coding-is-getting-expensive-use-more-open-models
- ainews-moonshot-kimi-k26-the-worlds-leading-open-model-refre
- alguma-expectativa-para-incluir-o-kimi-k26-na-cloud
- an-open-source-model-has-returned-to-1-on-the-3d-design-lead
- another-proof-point-for-the-open-weights-thesis-from-ramplab
- april-was-a-pretty-strong-month-for-llm-releases---gemma-4--
- bracesproul-changed-our-orgs-internal-model-in-fleet-from-so
- breaking-kimi-k3-by-kimimoonshot-is-1st-overall-on-3d-design
- cerebras-is-now-running-kimi-k26-a-trillion-parameter-model-
- claude-through-kimi-k26-on-ollama-security-concern
- dear-god-lol---new-kimi-model-is-a-fucking-beast-gpt-54-leve
- dell-nvidia
- heres-a-new-limited-time-top-up-reward-grab-bonus-vouchers-a
- i-remember-when-people-were-saying-its-useless-to-open-sourc
- icymi-kimi-k26-is-now-in-notion-its-the-first-open-weight-mo
GPT-5.6
Mentioned in 74 entries
- big-breaking-news-white-house-asks-openai-to-delay-gpt--56
- chatgpt-work-is-powered-by-gpt-56-gpt-56-makes-chatgpt-state
- claude-power-users-fable-5-is-the-best-codex-power-users-gpt
- everyone-should-pay-attention-to-the-training-timelines-in-t
- fable-feels-very-different-than-opus-gpt-56-feels-like-a-par
- finally-can-share-that-i-have-been-testing-gpt-56-early-and-
- good-new-first-sol-is-a-smart-efficient-and-a-significant-st
- gpt-56-found-optimizations-that-reduced-end-to-end-serving-c
- gpt-56-frontier-intelligence-that-scales-with-your-ambition
- gpt-56-is-a-major-step-forward-for-health-intelligence-acros
- gpt-56-is-here-but
- gpt-56-is-now-available-in-devin-on-fontiercode-11-extended-
- gpt-56-is-now-supported-in-hermes-agent-and-available-via-no
- gpt-56-is-now-the-preferred-model-in-microsoft-365-copilot
- gpt-56-is-now-the-preferred-model-in-microsoft-365-copilot-h
Astra
Mentioned in 70 entries
- 6-billion-tokens-for-65-is-it-as-good-as-astra-and-fable-at-
- a-lot-of-hype-around-openais-astra-model-here-on-my-timeline
- access-to-astra-is-rolling-out-over-the-coming-days-starting
- after-evaluating-one-of-our-upcoming-models-astra-were-treat
- agreed-one-of-the-biggest-bottlenecks-right-now-is-my-own-ab
- an-internal-version-of-astra-openais-next-major-model-family
- another-informal-nerd-benchmark-that-ais-had-struggled-with-
- anybody-use-frontier-models-like-astrafable-for-planningjudg
- as-we-prepare-to-release-astra-were-focused-on-making-increa
- astra-a-generalizable-report-generation-foundation-model-for
- astra-achieves-state-of-the-art-results-on-agents-last-exam-
- astra-is-a-genuine-improvement-but-not-even-off-trend-so-muc
- astra-is-one-of-the-best-models-weve-tested-at-parsing-compl
- astra-is-our-most-aligned-model-with-substantial-improvement
- astra-vs-fable-51-on-real-ml-tasks----tradeoffs-strengths-sh
Ideogram 4
Mentioned in 68 entries
- ainews-reve-2-and-ideogram-4-layouts-in-imagegen
- an-experiment-recreate-json-prompted-closed-model-image-in-i
- bounding-box-composition-control-in-ideogram-4-what-works-wh
- can-ideogram-4-do-512x512
- character-reference-sheets-with-ideogram-4-in-comfyui
- first-try-with-the-ideogram-4-prompt-builder-kj
- for-those-with-limited-disk-space-can-the-unconditional-mode
- fullrank-finetuning-ideogram4
- get-rid-of-image-blocked-by-safety-filter-in-ideogram-4
- getting-nice-results-from-ideogram-4
- how-are-you-guys-even-prompting-ideogram-4
- how-to-style-transfer-with-ideogram-4-training-free
- i-made-a-tool-to-turn-any-image-into-ideogram-json-prompt
- i-think-im-finally-feeling-ideogram-4
- i-upgraded-the-ideogram-4-prompt-builder-node-kjnodes-using-
Hermes Agent
Mentioned in 67 entries
- a-full-house-of-cool-people-learning-about-hermes-agent
- a-new-beginning-of-pc-starts-with-nvidiartxspark-superchargi
- a-trillion-tokens-a-day-and-200k-github-stars-very-proud-of-
- also-i-found-that-hermes-agent-nemotron-3-ultra-is-a-mighty-
- another-stealth-model-launch-space-bunny-alpha-is-now-free-o
- chatgpt-image-25-is-now-available-on-nous-portal-to-use-in-h
- check-out-openrouters-new-guide-on-getting-the-most-out-of-o
- computer-use-with-any-model-hermes-agent-trycua
- congrats-to-hermes-agent-and-the-whole-nousresearch-team
- cutting-edge-of-digital-art-with-hermes-agent
- deepseek-v4-flash-is-back-on-nous-portal-for-free-for-use-in
- everyone-interested-in-getting-the-most-out-of-hermes-agent-
- for-anyone-running-nousresearch-hermes-agent-locally-and-wis
- found-a-way-to-save-everyone-14-on-input-tokens-on-average-d
- give-our-early-preview-of-computer-use-with-any-model-a-try-
Mythos
Mentioned in 64 entries
- 1-agreed-w-scaling01-that-mythos-appears-to-be-better-gpt-55
- a-cybersecurity-harbinger-oracle-front-runs-ai-model-threat-
- agent-arena-code---very-good-result-preliminary-for-glm-and-
- and-also
- and-we-have-not-seen-mythos-or-whatever-openai-and-google-ar
- anthropic-didnt-just-add-guardrails-to-make-mythos-safer-the
- anthropic8217s-mythos-mess-is-only-getting-worse
- anthropic8217s-mythos-rollout-has-missed-americas-cybersecur
- anthropics-mythos-ai-model-sparks-fears-of-turbocharged-hack
- anthropics-mythos-has-been-accessed-by-a-small-group-of-unau
- anthropics-new-model-fable-will-silently-handicap-work-on-ll
- anthropics-own-internal-security-blows
- as-i-look-at-some-of-the-replies-which-are-startlingly-human
- breaking-anthropic-just-dropped-claude-fable-5this-is-mythos
- can-confirm-we-saw-a-strong-spike-in-growth-of-token-consump
LLMs
Mentioned in 59 entries
- a-unified-graph-language-model-for-multi-domain-multi-task-g
- absolutely-right-gary-llms-simply-dont-understand-causality-
- ai-generates-well-liked-but-templatic-empathic-responses
- ai-is-jagged-but-i-think-sometimes-it-is-easy-to-overly-focu
- audiorole-an-audio-dataset-for-character-role-playing-in-lar
- aura-internalizing-audio-understanding-into-llms-as-lora
- automated-rubrics-for-reliable-evaluation-of-medical-dialogu
- boosting-reinforcement-learning-with-verifiable-rewards-via-
- bootstrapping-post-training-signals-for-open-ended-tasks-via
- bridging-the-english-arabic-medical-knowledge-gap-targeted-l
- bridging-the-long-tail-gap-robust-retrieval-augmented-relati
- confrag-confidence-guided-retrieval-augmenting-generation
- domain-specific-data-generation-framework-for-rag-adaptation
- echoes-of-automation-the-increasing-use-of-llms-in-newsmakin
- evaluating-repository-level-software-documentation-via-quest
DeepSeek
Mentioned in 58 entries
- a-look-at-deepseeks-model-optimization-to-reduce-hbm-use-pot
- a-profile-of-yang-zhilin-founder-of-moonshot-ai-which-faced-
- and-now-a-new-deepseek-model-and-appears-to-be-fully-open-we
- anyone-wonder-why-americans-lag-so-far-behind-in-the-open-ll
- big-day-for-american-open-models-nemotron-3-ultra-is-now-the
- bro-i-need-you-to-focus
- btw-zai-ipoed-in-jan-at-hk120-a-share-when-i-first-met-lousz
- built-gpt-2-llama-3-and-deepseek-from-scratch-in-pytorch---o
- command-a-sets-a-new-high-for-coheres-machine-translation-ca
- deep-agents-v06-makes-harness-profiles-a-first-class-abstrac
- deepseek-and-other-multimodal-models-dont-accept-images-on-o
- deepseek-exposed-users-can-access-each-others-conversations-
- deepseek-harness-02---optional-bundle-architecture-windows-s
- deepseek-input-cache-price-drop-effective-immediately-the-pr
- detailed-review-and-guide-from-my-testing-of-local-ollama-se
CNN
Mentioned in 56 entries
- a-flow-rate-conserving-cnn-based-domain-decomposition-method
- a-hybrid-cnn-lstm-intrusion-detection-framework-for-cybersec
- a-lightweight-cnn-integrated-compact-convolutional-transform
- a-resource-efficient-hybrid-cnn-lstm-network-for-image-based
- accelerating-hevc-intra-partitioning-via-a-cnn-hierarchical-
- adjoint-inversion-reveals-holographic-superposition-and-dest
- adversarial-robustness-of-time-series-classification-for-cry
- aerial-wildfire-suppression-planning-with-a-hybrid-cnn-cellu
- analysis-of-the-shortcut-learning-and-clever-hans-effect-in-
- automated-brain-tumor-detection-in-mri-images-using-cnn-and-
- automatic-weld-seam-segmentation-for-industrial-quality-cont
- balancing-real-and-synthetic-data-for-cnn-based-masonry-crac
- classification-of-single-and-mixed-partial-discharges-under-
- cnn-based-multi-in-multi-out-model-for-efficient-spatiotempo
- cnns-transformers-hybrid-and-vision-language-models-for-skin
LSTM
Mentioned in 55 entries
- 21cmemuv3-a-hybrid-diffusion-lstm-emulator-of-21cmfast-summa
- a-comparison-of-traditional-machine-learning-algorithms-and-
- a-deep-reinforcement-learning-approach-to-automated-stock-tr
- a-hybrid-cnn-lstm-intrusion-detection-framework-for-cybersec
- a-hybrid-tucker-lstm-tensor-network-model-for-soc-prediction
- a-multi-timestep-lstm-ensemble-regressor-for-enhanced-short-
- a-novel-lstm-music-generator-based-on-the-fractional-time-fr
- a-resource-efficient-hybrid-cnn-lstm-network-for-image-based
- a-unified-particle-filter-lstm-for-data-driven-process-simul
- adaptive-online-learning-with-lstm-networks-for-energy-price
- aifl-a-global-daily-streamflow-forecasting-model-using-a-det
- air-quality-station-simulation-via-lstm-and-attention-based-
- an-objective-performance-evaluation-of-the-lstm-networks-in-
- automatic-field-of-view-adjustment-for-a-view-expansive-micr
- classification-of-public-opinion-on-the-free-nutritional-mea
Claude Opus 4.7
Mentioned in 55 entries
- a-quick-fix-if-you-saw-higher-rate-limit-usage-in-opus-47-to
- ai-coding-is-getting-expensive-use-more-open-models
- ainews-anthropic-claude-opus-47---literally-one-step-better-
- anthropic-launches-claude-opus-47-with-coding-visual-reasoni
- anthropic-releases-a-new-opus-model-amid-mythos-preview-buzz
- changes-in-the-system-prompt-between-claude-opus-46-and-47
- claude-code-spend-had-gotten-to-1095m-runrate-peak-at-semian
- claude-opus-47-fast-mode-is-now-available-in-windsurf-full-c
- claude-opus-47-is-now-available-as-an-agent-preview-inside-o
- claude-opus-47-is-now-available-in-cursor-weve-found-it-to-b
- claude-opus-47-is-now-available-in-windsurf-20-anthropic-has
- claude-opus-47-is-now-the-default-orchestration-model-poweri
- claude-opus-47-just-implemented-an-alphazero-style-self-play
- claude-opus-47-on-ai-gateway
- claude-opus-47-on-vertex-ai
Qwen 3.6
Mentioned in 54 entries
- 180-toks-generation-on-a-4090-with-qwen-36-if-youre-on-a-409
- anyone-else-amped-up-over-qwen-38
- april-was-a-pretty-strong-month-for-llm-releases---gemma-4--
- basic-coding-system-demo-via-ollama---quick-head-to-head-gem
- citizen-kane-intro-but-its-all-ai---qwen-36-ltx-23
- codex-cli-local-qwen36-on-rx-9070-xt-16gb-egpu
- como-de-bueno-es-qwen-36-de-ollama
- compared-qwen36-qwen3-coder-and-deepseek-coder-on-three-codi
- current-smallest-usable-coding-model
- deepseek-v4-flash-284b-on-53gb-of-memory
- does-anyone-know-the-parser-and-render-for-qwen36-in-ollama-
- fixed-jinja-chat-template-for-qwen-35-36-and-the-new-38-rele
- gemma-4-adoption-numbers-outpacing-qwen-3536-for-the-same-si
- harness-am-i-doing-something-wrong-or-are-my-expectations-un
- how-does-qwen-38-27b-compare-on-low-thinking-mode-to-the-old
DeepSeek V4 Flash
Mentioned in 54 entries
- 4xradeon-ai-pro-r9700-people-how-are-your-benchmarks
- another-day-and-another-full-frontier-model-running-on-your-
- anyone-using-koreas-solar-open-2-250b-15b-model
- b10254
- day-0-support-for-deepseekai-v4-pro-and-flash-on-vllm-a-new-
- deepseek-releases-its-new-flagship-models-v4-pro-and-v4-flas
- deepseek-rolls-out-the-official-v4-flash-api-in-public-beta-
- deepseek-says-it-plans-to-implement-substantial-price-increa
- deepseek-unveils-an-experimental-multimodal-version-of-its-v
- deepseek-v4-ai-beats-billion-dollar-systemsfor-free
- deepseek-v4-flash-0731-uncensored-jailbreak-pt2
- deepseek-v4-flash-ga-ranks-the-same-as-sonnet-5-and-grok-45-
- deepseek-v4-flash-has-a-nice-bump-in-capability
- deepseek-v4-flash-is-90-cheaper-than-gpt-54-mini-and-70-chea
- deepseek-v4-flash-is-back-on-nous-portal-for-free-for-use-in
LTX-2.3
Mentioned in 52 entries
- a-compilation-of-the-open-source-loras-for-ltx-23---released
- another-video-from-ltx-23-distilled
- any-way-to-get-ltx23-i2v-to-work-on-32gb-integrated-chip-gpu
- audiocpp-release-05-dramabox-expressive-tts-confucius4-cross
- chrono-trigger-remake-concept-made-in-ltx-23
- comfyui-dramabox-now-supports-loras-and-voice-clone-studio-d
- does-ltx-23-have-good-motion-transfer
- dramabox---most-expressive-voice-model-ever-based-on-ltx-23
- editanything-ic-lora---ltx-23
- grok-and-ltx-23-is-the-best-combo-made-my-own-trailer
- has-everyone-moved-onto-ltx-23-then
- help-with-ltx-23-workflow
- how-to-use-ltx-director---a-free-tool-for-creating-advanced-
- im-trying-out-ltx-23-as-well
- is-there-any-v2v-workflows-for-ltx23
SAM
Mentioned in 52 entries
- a-3d-sam-based-progressive-prompting-framework-for-multi-tas
- ahcq-sam-toward-accurate-and-hardware-compatible-post-traini
- amodal-sam-a-unified-amodal-segmentation-framework-with-gene
- an-edge-aware-prompt-enhanced-sam-for-ultrasound-image-segme
- boxes2pixels-learning-defect-segmentation-from-noisy-sam-mas
- bridging-semantics-and-geometry-a-decoupled-lvlm-sam-framewo
- bringing-sam-to-new-heights-leveraging-elevation-data-for-tr
- can-frozen-hyperspherical-features-guide-the-selection-of-ps
- clip-guided-label-free-discriminative-region-scoring-for-fin
- crack-adversarial-attacks-on-cross-model-consistency-in-coll
- crc-sam-sam-based-multi-modal-segmentation-and-quantificatio
- diffusam-diffusion-based-prompt-free-sam2-for-few-shot-and-s
- dont-waste-sam
- et-sam-efficient-point-prompt-prediction-in-sam-for-unified-
- exploring-sam-supervision-for-fine-grained-uav-target-segmen
Qwen3.8-Flash-Next
Mentioned in 52 entries
- 22-hours-to-go-for-qwen38-flash-there-is-no-sleep-in-local-a
- amazing-thanks-vllmproject-for-getting-qwen38-flash-next-run
- at-a-1m-token-context-length-qsas-attention-kernel-is-up-to-
- atomicchatqwen38-flash-next-gguf-is-really-good
- b10660
- big-thanks-to-sglproject-for-the-day-0-support-qwen38-flash-
- built-this-yesterday-with-qwen38-flash-next-nvfp4-262k-conte
- deepseek-v4-flash-vision-exp-is-amazing-at-creating-game-wor
- deepseek-v4-flash-vision-q8-vs-qwen38-flash-next-q8
- experiment-with-qwen38-flash-next-on-nvidia-gb300-nvl72-for-
- faster-than-light-in-air-8-22-tgs-qwen38-flash-next-q4q4ish-
- llamacpp-ngram-on-ramssd
- m5-ultra-96gb-vs-m5-max-128gb-is-2x-bandwidth-worth-losing-3
- on-the-design-of-qwen38-next-architecture-evaluation-efficie
- optimal-125-bit-quantization-of-qwen38-flash-next
Qwen 3.8
Mentioned in 52 entries
- 1bit-qwen-38-24t-a95b-unsloth-iq1s-medium-reasoning
- 1x32gb-v100-vs-2x16gb-v100-vs-5060ti-16gb-for-qwen-38
- 4-5-days-replacing-claude-w-qwen-38-next
- a-27b-model-beating-latest-frontier-models-was-not-on-my-202
- amd-setup-is-fast-with-200k-ctx-with-qwen-38-i-didnt-underst
- anybody-use-frontier-models-like-astrafable-for-planningjudg
- anyone-else-amped-up-over-qwen-38
- anyone-sitting-on-a-lot-of-slow-system-memory-and-a-modest-g
- been-tweaking-my-qwen-38-setup-up-to-45-steady-tps-at-8bit-q
- best-hardware-for-qwen-38
- company-approved-128gb-mac-for-research-proposal-best-model
- confirmed-bolting-q8-ngram-into-iq4-qwen-no-speed-degradatio
- deepseek-harnness---why-is-feels-better
- dual-7900-xtx---some-guy-made-a-pretty-optimized-fork-of-lam
- fable-5-refuses-to-touch-qwen-deployments
VLA
Mentioned in 50 entries
- abot-m0-vla-foundation-model-for-robotic-manipulation-with-a
- abot-m05-unified-mobility-and-manipulation-world-action-mode
- actionground-training-free-runtime-refinement-of-frozen-vla-
- adaptive-action-chunking-at-inference-time-for-vision-langua
- ai-lecun-llm-metametallm-pilled
- anatomy-of-a-closed-loop-collapse-a-causal-case-study-of-a-c
- anycamvla-zero-shot-camera-adaptation-for-viewpoint-robust-v
- bridging-the-morphology-gap-adapting-vla-models-to-dexterous
- cloak-zero-shot-cross-embodiment-manipulation-by-masking-the
- crosssafe-towards-cross-embodiment-latent-safety-filters
- crosstracer-cross-embodiment-navigation-via-vla-model-reason
- dreamavoid-critical-phase-test-time-dreaming-to-avoid-failur
- dynawm-a-base-vla-guided-world-foundation-model-for-moving-o
- efficient-long-horizon-vision-language-action-models-via-sta
- eximo-vlm-guided-exploration-of-vla-policies
LLM
Mentioned in 50 entries
- a-gan-and-llm-driven-data-augmentation-framework-for-dynamic
- ai-lecun-llm-metametallm-pilled
- ai-speech-to-speech
- aligning-language-models-for-lyric-to-melody-generation-with
- an-empirical-study-of-lora-based-fine-tuning-of-large-langua
- analyzing-multimodal-interaction-strategies-for-llm-assisted
- attention-guided-reliability-scaling-for-contrastive-decodin
- bert-as-a-judge-a-robust-alternative-to-lexical-methods-for-
- beyond-surface-judgments-human-grounded-risk-evaluation-of-l
- breaking-the-self-confirming-loop-diagnosing-and-mitigating-
- can-legal-ai-know-when-it-is-wrong-and-do-students-know-when
- can-we-trust-llms-logic-quantifying-uncertainty-coherence-an
- cofi-pgma-counterfactual-policy-gradients-under-filtered-fee
- cuda-l2-surpassing-cublas-performance-for-matrix-multiplicat
- daily-and-weekly-periodicity-in-large-language-model-perform
Claude Opus 4.8
Mentioned in 49 entries
- a-27b-agent-just-beat-claude-opus-48-and-gpt-55-on-held-out-
- all-oneshots-from-kimi-k3-looks-better-than-opus48
- anthropic-launches-claude-science-an-ai-workbench-that-uses-
- anthropic-launches-claude-sonnet-5-saying-it-nears-opus-48-p
- anthropic-says-claude-fable-5-uses-conservative-safety-class
- as-anthropic-launches-claude-opus-48-it-raises-65b-in-new-fu
- b10517
- claude-in-microsoft-foundry-is-now-generally-available-hoste
- claude-opus-48-a-modest-but-tangible-improvement
- claude-opus-48-and-sonnet-5-seem-worse-at-tool-calls-than-ol
- claude-opus-48-benchmark-numbers-vs-gpt-55-are-kinda-concern
- claude-opus-48-is-now-available-for-max-subscribers-on-perpl
- claude-opus-48-is-now-available-in-cursor-on-cursorbench-its
- claude-opus-48-is-now-available-in-windsurf-and-devin-cli
- claude-opus-48-is-now-available-on-aws
BERT
Mentioned in 47 entries
- a-bertology-view-of-llm-orchestrations-token--and-layer-sele
- a-comparative-evaluation-of-structural-topic-models-and-bert
- a-fine-tuned-bert-classifier-for-personal-letter-titles-in-l
- a-mechanistic-study-of-ai-text-detection-neurons-in-frozen-b
- assessing-reliability-of-bert-based-models-on-question-answe
- bamibert-a-new-bert-based-language-model-for-vietnamese
- bert-as-a-judge-a-robust-alternative-to-lexical-methods-for-
- bert-based-models-vs-large-language-models-for-low-resource-
- bert4dti-bert-based-model-for-predicting-drug-protein-intera
- cnm-bert-a-drop-in-structural-embedding-for-chinese-characte
- comparing-bert-sentence-pair-classification-and-few-shot-llm
- construction-of-historical-knowledge-graphs-based-on-bert-an
- continuous-behavioral-authentication-via-multi-expert-bert-l
- do-bert-embeddings-encode-narrative-dimensions-a-token-level
- efficient-int8-inference-of-small-nlp-models-on-server-cpus-
Whisper
Mentioned in 46 entries
- a-semi-supervised-framework-for-speech-confidence-detection-
- a-speech-corpus-for-mizo-automatic-speech-recognition-whispe
- a-study-of-asr-adaptation-and-representation-dimensionality-
- ai-is-watching-a-film-via-marlinvisuals-whisperaudio-and-pal
- any-better-local-alternative-to-whisperer
- architecture-advice-real-time-pipeline-for-youtube-audio---w
- askd-whisper-adaptive-self-knowledge-distillation-for-effici
- automatic-lyric-transcription-for-greek-songs-scaling-and-ta
- b10672
- baldwhisper-faster-whisper-with-head-shearing-and-layer-merg
- baltivoice-a-speech-corpus-and-fine-tuned-whisper-asr-system
- bangla-whisperdiar-fine-tuning-whisper-and-pyannote-for-bang
- banglaturn-a-benchmark-and-whisper-based-model-for-end-of-tu
- best-current-methods-for-finetuning-whisper-on-domain-specif
- buzzasr-a-swarm-of-100-monolingual-speech-recognition-models
Opus 4.8
Mentioned in 45 entries
- a-red-team-study-of-anthropic-fable-5-opus-48-models
- a-tangible-comparison-of-glm-performance-stacked-against-opu
- after-using-glm-52-for-a-day-im-surprised-by-how-often-it-fe
- ainews-anthropic-raises-965b-series-h-releases-opus-48-and-d
- all-currently-popular-local-models-in-one-table-opus-48-resu
- announcing-glm-arena-a-series-of-tests-infographics-svgs-sit
- anthropic-says-some-routine-tasks-like-coding-and-debugging-
- anyone-else-still-using-opus-46-over-48
- breaking-anthropic-just-dropped-opus-48and-it-is-a-monster-w
- but-but-i-thought-it-was-magic
- cant-merge-claude-max-subscription-opus-48-with-zai-glm-521m
- cognition-releases-swe-17-trained-from-kimi-k27-and-availabl
- did-something-happen-to-opus-48-today-first-time-ive-seen-th
- glm-52-is-5x-cheaper-than-opus-48-and-11x-than-fable-5-yet-i
- glm-52-is-already-top-10-on-openrouter-this-week-sitting-rig
GLM-5.1
Mentioned in 44 entries
- 202618glm-51linux-202618linuxaiglm-51swe
- 754b-parameters-151tb-on-hugging-face
- a-lot-of-you-are-asking-about-the-cloud-tag-heres-the-deal-g
- agent-glm-51-swe-bench-pro-1584-3-agent-8-50-app-linux
- aihttpzaizhipu-aiaiglm-51-swe-bench-pro58413gpt
- april-was-a-pretty-strong-month-for-llm-releases---gemma-4--
- check-out-the-glm-51-first-impressions-with-peter-on-our-you
- come-check-out-glm-51-in-code-arena-for-agentic-web-developm
- fantastic-to-see-glm-being-applied-to-such-fresh-dynamic-sce
- glm-51-by-zaiorg-is-now-3-in-code-arena---surpassing-gemini-
- glm-51-by-zaiorg-just-launched-in-the-text-arena-and-is-now-
- glm-51-feels-very-very-very-slow-on-ollama-cloud
- glm-51-from-zaiorg-is-now-live-on-orcarouter-1-open-source-m
- glm-51-gives-teams-a-stronger-model-for-coding-tool-use-and-
- glm-51-is-coming-httpshuggingfacecozai-orgglm-51-coding-is-t
Vision-Language Models
Mentioned in 44 entries
- act2see-emergent-active-visual-perception-for-video-reasonin
- adaboosting-text-prompts-for-vision-language-models
- agmark-attention-guided-dynamic-watermarking-for-large-visio
- aligning-what-vision-language-models-see-and-perceive-with-a
- analysis-by-proxy-localization-signals-in-vlms-operating-as-
- anchorgui-asymmetric-memory-for-dual-scale-learning-in-gui-n
- apt-action-expert-pretraining-improves-instruction-generaliz
- be-faithful-when-response-returning-fluent-and-grounded-answ
- can-linguistic-reasoning-vectors-enhance-multimodal-reasonin
- dics-exploring-data-intrinsic-consistency-for-visual-instruc
- disentangling-algorithmic-bias-from-archival-artifacts-a-con
- does-forgetting-transfer-across-modalities-a-real-world-benc
- dual-stream-semantic-guidance-with-prototype-anchor-calibrat
- encore-entropy-guided-cropping-and-attention-regularization-
- geoworld-vlm-geometry-from-world-models-for-vision-language-
DeepSeek V4
Mentioned in 44 entries
- another-proof-point-for-the-open-weights-thesis-from-ramplab
- b10021
- b10206
- b10232
- b10593
- b11020
- build-with-deepseek-v4-using-nvidia-blackwell-and-gpu-accele
- can-i-run-deepseek-v4-on-my-rig
- chinas-deepseek-previews-new-ai-model-a-year-after-jolting-u
- deepseek-details-dspark-a-speculative-decoding-framework-for
- deepseek-has-began-grayscale-testing-for-deepseek-with-visio
- deepseek-open-sources-v4-large-language-model-series
- deepseek-v4---almost-on-the-frontier-a-fraction-of-the-price
- deepseek-v4-ai-beats-billion-dollar-systemsfor-free
- deepseek-v4-by-deepseekai-just-dropped-sglang-is-ready-on-da
Opus 4.7
Mentioned in 43 entries
- a-major-lesson-to-take-away-from-opus-47-is-that-while-there
- among-all-the-model-release-furor-important-to-note-that-peo
- anthropic-says-opus-47-hits-806-on-document-reasoning-up-fro
- anthropic-says-the-models-that-breached-three-companies-incl
- april-2026-newsletter
- breaking-the-results-are-in-for-slides-arena-anthropicai-and
- but-but-i-thought-it-was-magic
- chinese-ai-developer-minimax-launches-m3-a-new-coding-model-
- cursor-composer-25s-is-318x-cheaper-than-opus-47-in-claude-c
- cursors-new-composer-25-takes-third-on-the-artificial-analys
- dogfooding-opus-47-the-last-few-weeks-ive-been-feeling-incre
- for-anyone-curious-this-was-the-result-of-many-experiments-b
- great-experiment-testing-how-good-ais-are-getting-at-very-am
- grok-1-in-law
- happy-coding-opus-47-is-a-significant-step-up-to-get-the-mos
GPT-4
Mentioned in 41 entries
- a-comprehensive-survey-of-agents-for-computer-use-foundation
- a-lot-of-papers-coming-out-are-still-focused-on-gpt-4-but-yo
- a-tool-augmented-gpt-4-chatbot-for-real-time-repository-data
- ai-models-are-terrible-at-betting-on-soccerespecially-xai-gr
- and-microsoft-rolled-out-gpt-4-first
- and-yes-our-experiments-used-a-mix-of-gpt-4-gpt-4o-publishin
- are-gui-agents-focused-enough-automated-distraction-via-sema
- assessing-capabilities-of-large-language-models-in-social-me
- claude-code-still-not-agi-but-biggest-advance-since-gpt-4-is
- commander-gpt-dividing-and-routing-for-multimodal-sarcasm-de
- deep-learning-hit-a-wall-neurosymbolic-ai-rescued-it
- from-gpt-3-to-gpt-5-mapping-their-capabilities-scope-limitat
- from-gpt-4-to-gpt-6-sparks-was-a-remarkably-prescient-paper-
- given-how-much-of-the-original-bottle-of-water-per-generated
- given-today-it-is-surprising-how-daring-microsoft-google-wer
GPT-5
Mentioned in 41 entries
- a-lot-of-papers-coming-out-are-still-focused-on-gpt-4-but-yo
- aihttpzaizhipu-aiaiglm-51-swe-bench-pro58413gpt
- azure-content-understanding-gpt-5-series-guide-model-selecti
- by-end-of-next-year-end-of-year-after-for-16gb-vram-do-folk-
- chatgpt-is-struggling
- could-a-sequence-of-poetic-gibberish-manipulate-the-worlds-m
- dont-think-i-have-seen-more-misconceptions-around-one-model-
- entry-level-guide-to-the-use-of-large-language-models-for-me
- fable-feels-very-different-than-opus-gpt-56-feels-like-a-par
- from-sync-apis-to-support-for-the-gpt-5-model-series-and-age
- gpt-55
- gpt-56-sol-terra-and-luna-are-now-available-in-cursor-on-cur
- gpt-6-is-pretty-good-tbh
- how-gpt-5-helped-immunologist-derya-unutmaz-solve-a-3-year-o
- i-called-both-missteps-and-thats-part-of-why-openai-and-thei
Gemma
Mentioned in 37 entries
- agent-harnesses-have-an-expiration-date
- alibabas-open-weight-ai-models-accumulated-3-billion-global-
- announcing-the-gemma-challenge-google-hugging-face-and-the-o
- awesome-gemma-is-live-on-github-a-list-of-awesome-gemma-reso
- can-gemma-and-qwen-models-catch-hallucinations-by-looking-at
- congratulations-to-the-whole-googledeepmind-gemma-team-gemma
- deepseek-details-dspark-a-speculative-decoding-framework-for
- diffusiongemma-4x-faster-text-generation
- distilledgemma-balanced-efficiency-accuracy-for-person-place
- drummers-artemis-31b-v1-and-v11---coming-back-with-a-bang
- earth-day-at-googlecloudnext-im-demoing-a-sustainability-age
- enabling-intrinsic-reasoning-over-dense-geospatial-embedding
- event-with-googlegemma-next-week-hang-out-with-the-fellow-bu
- folks-gemma-is-a-good-model-very-useful-for-some-purposes-it
- google-deepmind-releases-diffusiongemma-a-model-that-runs-lo
Segment Anything Model
Mentioned in 37 entries
- 3d-segment-anything-model-with-visual-mamba-for-diagnosing-p
- ad-sam-adapting-the-segment-anything-model-for-semantic-segm
- ahcq-sam-toward-accurate-and-hardware-compatible-post-traini
- automatically-redact-pii-in-images-with-amazon-nova
- car-sam-cross-attention-reconstruction-for-post-training-qua
- confidence-calibrating-regularization-for-robust-brain-mri-s
- contour-field-based-elliptical-shape-prior-for-the-segment-a
- cross-cascaded-distillation-and-dual-constraint-grounding-fo
- diffusam-diffusion-based-prompt-free-sam2-for-few-shot-and-s
- dont-waste-sam
- fan-lora-a-fourier-adaptive-nonlinear-low-rank-adaptor-for-m
- frequency-and-edge-guided-segment-anything-model-for-remote-
- geosam-lite-a-lightweight-foundation-model-for-onboard-remot
- i-tested-huggingface-ml-intern-given-the-prompt-fine-tune-a-
- lighting-aware-unified-model-for-instance-segmentation
Nemotron 3 Ultra
Mentioned in 37 entries
- ainews-nvidia-cosmos-3-nemotron-3-ultra-and-rtx-spark
- also-i-found-that-hermes-agent-nemotron-3-ultra-is-a-mighty-
- and-another-open-weight-release-nemotron-3-ultra-has-an-ultr
- big-day-for-american-open-models-nemotron-3-ultra-is-now-the
- box-agent-uses-langchains-deep-agents-harness-to-bring-speci
- busy-couple-days-ahead-today-nvidia-nemotron-3-ultra-support
- creating-the-nvidia-nemotron-3-ultra-nvfp4-checkpoint-with-n
- dcode-nemotron-3-ultra-openshell-never-been-easier-to-own-yo
- deepagent-for-competitive-analysis
- fireworks-training-platform-keeps-expanding-leading-us-open-
- harrison-chase-hwchase17-nemotron-3-ultra-hit-86-claude-opus
- how-full-stack-nim-optimizations-deliver-25x-more-users-on-n
- i-benchmarked-9-open-models-on-spotting-fake-sources-during-
- icymi-langchain-deep-agents-on-nvidia-nemotron-3-ultra-front
- introducing-signals-of-ai---deepagent-powered-by---nebiustf-
Fable
Mentioned in 36 entries
- 6-billion-tokens-for-65-is-it-as-good-as-astra-and-fable-at-
- absurd-claim-the-distilled-model-outperforms-the-originals
- ainews-the-field-guide-to-fable
- anybody-use-frontier-models-like-astrafable-for-planningjudg
- benched
- big-gains-in-potatomonetbench-from-opus-45-to-fable-there-ar
- claude-opus-50-identified-the-ip-sharing-route-and-declined-
- even-though-i-made-this-graph-it-is-also-kind-of-wrong-fable
- fable-feels-very-different-than-opus-gpt-56-feels-like-a-par
- fable-gets-another-bump
- fable-the-indomitable-model-from-anthropic-is-coming-back-th
- fables-judgement
- fugu-stands-shoulder-to-shoulder-with-leading-models-like-fa
- generate-a-fake-but-believable-witty-churchill-insult-at-a-p
- i-found-myself-switching-between-fable-and-sol-depending-on-
Grok 4.3
Mentioned in 36 entries
- an-analysis-of-gpt-55-gemini-31-pro-grok-43-gabs-arya-and-ot
- both-gemini-and-grok-are-underrated-i-used-to-champion-gemin
- grok-1-in-law
- grok-43
- grok-43---excellent-intelligence-per-unit-cost
- grok-43-beta-just-went-insane-with-front-end-design-i-gave-i
- grok-43-beta-passes-the-caitlyn-jenner-ai-test-chatgpt-would
- grok-43-betas-front-end-capabilities-are-totally-insane-i-ju
- grok-43-can-take-in-video-and-extract-audio-files
- grok-43-is-better-than-people-think
- grok-43-is-next-level-ive-been-using-it-quite-a-bit-for-ever
- grok-43-is-still-an-early-beta-that-will-improve-almost-ever
- grok-43-just-built-this-entire-game-with-just-a-single-promp
- grok-43-just-leveled-up-it-can-create-slides
- grok-43-now-has-native-latex-compilation-built-into-grok-fil
Claude Opus 5
Mentioned in 36 entries
- a-reminder-this-is-20-off-via-the-nous-portal-3
- am-i-supposed-to-interpret-that-its-better-than-fable-5-from
- amazon-bedrock-expands-claude-model-availability-to-in-count
- anthropic-launches-claude-opus-5-which-it-says-comes-close-t
- anthropic-launches-claude-opus-5-with-efficiency-safety-impr
- artificial-analysis-coding-agent-index-gpt-6-astra-scored-67
- b10417
- breaking-grok-45-ranked-1-on-laurenbench-with-a-score-of-569
- breaking-grok-46-just-tied-for-1-on-the-artificial-analysis-
- claude-opus-5-one-shotted-this-game-everything-you-see-in-th
- deepseek-v4-flash-on-slopcodebench
- does-a-tool-result-carry-more-authority-than-plain-text-thre
- google-gives-all-of-its-engineers-claude-opus-5-access-via-a
- google-launches-gemini-38-flash-cyber-for-partners-in-its-ne
- gpt-6-astra-scores-627-on-arc-agi-3-with-the-standard-harnes
ChatGPT Images 2.0
Mentioned in 35 entries
- access-gpt-image-20-natively-in-hermes-agent-update-now-to-g
- apparently-the-new-image-model-is-good-in-morphing-images
- aspect-ratios-resolution-in-chatgpt-images-20-demonstrated-b
- chatgpt-images-20-handles-text-so-perfectly-that-gemini-omni
- chatgpt-images-20-nails-photo-restoration-instead-of-just-re
- chats-new-image-generator-is-the-first-to-beat-my-robot-hair
- flexible-aspect-ratios-chatgpt-images-20-supports-aspect-rat
- gpt-image-2-is-really-good
- gpt-image-2-is-superb
- gpt-image-guides-20---2025-vs-2026
- i-came-up-with-a-somewhat-foolish-new-benchmark-for-testing-
- images-20-is-insane
- important-it-has-been-confirmed-that-chatgpt-images-20-added
- increased-free-image-generation-and-files
- introducing-chatgpt-images-20-a-state-of-the-art-image-model
DeepSeek V4 Pro
Mentioned in 35 entries
- 4444
- a-research-team-that-includes-huawei-says-it-successfully-us
- ainews-deepseek-v4-pro-16t-a49b-and-flash-284b-a13b-base-and
- an-evaluation-by-nists-caisi-says-deepseek-v4-pro-lags-behin
- anyone-else-finding-deepseek-v4-pro-unbearably-slow-on-ollam
- bro-wtf-qwen-lab-cooked-with-qwen-38-27b-its-so-fucking-good
- day-0-support-for-deepseekai-v4-pro-and-flash-on-vllm-a-new-
- deepseek-releases-its-new-flagship-models-v4-pro-and-v4-flas
- deepseek-says-it-will-lower-v4-pro-api-prices-by-75-to-04351
- deepseek-v4-ai-beats-billion-dollar-systemsfor-free
- deepseek-v4-flash-0731-is-live-on-fireworks-day-zero-deepsee
- deepseek-v4-pro-0813-is-live-on-together-ai-deepseeks-flagsh
- deepseek-v4-pro-brings-long-context-reasoning-and-sota-codin
- deepseek-v4-pro-costs-1741m-input-tokens-and-3481m-output-to
- deepseek-v4-pro-has-16t-total-parameters-its-largest-model-b
U-Net
Mentioned in 33 entries
- a-conditional-u-net-pipeline-with-pre--and-post-processing-f
- a-deep-u-net-framework-for-flood-hazard-mapping-using-hydrau
- adaptive-dual-residual-u-net-with-attention-gate-and-multisc
- attention-mamba-a-mamba-enhanced-multi-scale-parallel-infere
- automated-palynological-analysis-system-integrating-deep-met
- damamba-unet3d-a-parameter-efficient-mamba-state-space-u-net
- determining-vertical-displacement-of-agricultural-areas-usin
- does-your-vit-still-need-u-net-for-segmentation
- ea-liteunet-an-edge-adaptive-and-resource-efficient-u-net-fo
- energy-efficient-cnn-acceleration-with-msdf-digit-serial-ari
- epra-u-net-an-efficient-pyramid-residual-attention-framework
- farmland-extent-and-visible-boundary-mapping-from-1-m-naip-i
- from-convolution-to-transformer-a-comparative-study-of-u-net
- gaze-into-the-details-locality-sensitive-enhancement-for-oct
- hq-unet-a-hybrid-quantum-classical-u-net-with-a-quantum-bott
YOLO
Mentioned in 33 entries
- a-multi-dataset-benchmark-of-yolo-based-weed-detection-in-pr
- application-research-of-a-deep-learning-model-integrating-cy
- beyond-mamba-enhancing-state-space-models-with-deformable-di
- bladeyolo-wind-turbine-blade-defect-detection-with-limited-a
- broncholumen-analysis-of-recent-yolo-based-architectures-for
- cf-yolo-context-aware-feature-refinement-for-camouflaged-ind
- classroom-behavior-monitoring-with-yolo-an-empirical-study-i
- cwt-enhanced-vibration-sensing-with-spatial-fault-localizati
- depth-aware-pothole-detection-using-yolo-and-rt-detr-at-the-
- dronescan-yolo-redundancy-aware-lightweight-detection-for-ti
- ev-yolo-uncertainty-aware-object-detection-via-a-unified-evi
- fursee-hybrid-yolo-dinov3-framework-for-fursuit-identity-ret
- gsa-yolo-a-high-efficiency-framework-via-structured-sparsity
- hyolo-an-intelligent-iot-based-object-detection-system-using
- impact-of-dataset-composition-on-embedded-real-time-uav-wild
JEPA
Mentioned in 32 entries
- a-lightweight-library-for-energy-based-joint-embedding-predi
- a-slightly-improved-dvd-jepa-demo-p
- anisotropic-representations-improve-planning-in-jepa-world-m
- arc-bench-closed-loop-replanning-masks-broken-action-ranking
- arti-jepa-adapting-video-world-model-to-real-time-mri-of-the
- asleep-at-the-wheel-jepas-limitations-in-evaluating-novel-dr
- behavioral-monitoring-of-jepa-world-models-with-jacobian-cen
- can-regularization-based-jepa-eg-sigreg-scale-and-compete-wi
- cf-jepa-improving-robustness-of-jepa-world-models-via-contro
- cojepa-combining-contrastive-learning-and-jepa-for-global-lo
- contrast-encodes-inductive-bias-separating-slow-noise-from-d
- demo-jepa-joint-embedding-predictive-architecture-for-one-sh
- does-latent-planning-survive-point-clouds-action-conditioned
- dvd-jepa-an-open-source-fully-reproducible-jepa-world-model-
- earlier-this-year-yann-lecun-left-meta-because-mark-zuckerbe
DINO
Mentioned in 32 entries
- a2dinov3-rethinking-multi-modal-object-detection-via-sociali
- adadino-pair-aware-in-backbone-adaptation-of-frozen-dino-for
- affectflow-dino-uncertainty-aware-multi-task-affect-estimati
- agricountdino-parameter-efficient-exemplar-guided-counting-a
- can-regularization-based-jepa-eg-sigreg-scale-and-compete-wi
- crack-adversarial-attacks-on-cross-model-consistency-in-coll
- d3gs-depth-dino-and-rgb-diffusion-co-guided-3d-gaussian-spla
- ddms-discriminative-distillation-of-multi-view-foundational-
- deploy-dino-with-many-to-many-association
- dino-3dra-leveraging-2d-foundation-model-semantics-for-3d-ce
- dino-eats-clip-adapting-beyond-knowns-for-open-set-3d-object
- dino-soars-dinov3-for-open-vocabulary-semantic-segmentation-
- dinorade-full-spectral-radar-camera-fusion-with-vision-found
- dinov3-mil-per-kidney-multi-label-tumour-and-cyst-detection-
- driving-with-dino-vision-foundation-features-as-a-unified-br
Claude Opus
Mentioned in 32 entries
- 15-when-we-saw-our-reducer-llm-judge-component-in-maestro-ou
- b11355
- both-are-great-models-and-neither-wins-everywhere-i-use-both
- browser-extension-for-hermes-agent-nousresearch-hermes-lives
- claude-opus---glm-52-in-under-10-seconds-dcode-isnt-just-pro
- claudes-voice-mode-is-now-available-for-opus-and-sonnet
- copilot-just-9xd-sonnet-and-27xd-opus-and-teams-have-no-idea
- dcode-does-a-bunch-of-work-to-standardize-message-format-mak
- glm-52-is-great-at-design-opus-level-imo-i-am-also-starting-
- harrison-chase-hwchase17-nemotron-3-ultra-hit-86-claude-opus
- i-ran-glm-52-with-opencode-harness-against-claude-opus-this-
- i-spent-the-night-testing-open-source-coding-models-against-
- i-was-told-by-anthropic-that-they-are-looking-at-ways-of-fix
- interesting-comparison-of-grok-opus-1m-context-window-coming
- introducing-firerouter-with-opus-our-first-router-model-rout
Kolmogorov-Arnold Networks
Mentioned in 32 entries
- abfr-kan-kolmogorov-arnold-networks-for-functional-brain-ana
- adaptive-rbf-kan-a-comparative-evaluation-of-dynamic-shape-p
- an-embedded-risc-v-evaluation-of-kolmogorov--arnold-networks
- autocorrelation-reintroduces-spectral-bias-in-kans-for-time-
- delta-position-estimation-based-imu-odometry-a-comparison-of
- draining-fictitious-knots-restoring-distance-awareness-guara
- er-kans-efficient-and-robust-kolmogorov-arnold-networks-for-
- fisher-simplicity-in-kolmogorov-arnold-networks-and-multilay
- funccode-compressing-kolmogorov--arnold-networks-in-function
- geometry-aware-r-structured-kolmogorov-arnold-networks
- geometry-constrained-kolmogorov-arnold-networks-learning-edg
- gs-kan-parameter-efficient-kolmogorov-arnold-networks-via-sp
- hybrid-quantum-inspired-kolmogorov-arnold-networks-for-priva
- hydra-hyperbolic-dynamic-representation-architecture-for-kol
- implicit-spatial-frequency-fusion-of-hyperspectral-and-lidar
GPT-5.6 Luna
Mentioned in 32 entries
- advancing-the-price-performance-frontier-with-gpt56
- ainews-openai-gpt-56-sol-terra-luna-restricted-to-trusted-pa
- all-models-are-currently-20-discounted-in-portal-other-than-
- an-underrated-part-of-glm-53-is-its-low-hallucination-rate-c
- deepseek-v4-flash-0731-vs-gpt-56-luna-on-deepswe-cost-and-co
- get-started-with-openai-gpt-56-sol-terra-and-luna-on-amazon-
- gpt-56-is-a-major-step-forward-for-health-intelligence-acros
- gpt-56-sol-costs-5-per-1m-input-tokens-and-30-per-1m-output-
- gpt-6-luna-is-half-the-price-of-56-luna-which-was-already-an
- imajev-4b-i-spent-15-days-fine-tuning-a-4b-model-to-make-bus
- in-addition-to-the-upgrade-in-intelligence-with-gpt-56-luna-
- introducing-a-limited-preview-of-gpt-56-sol-our-next-generat
- introducing-explicit-prompt-caching-for-openai-gpt-56-models
- llm-032rc2
- maximize-making-minimize-cost-replit-free-mode-powered-by-op
Anima
Mentioned in 31 entries
- anima-can-edit-images-and-this-is-possible-in-two-different-
- anima-datasetissues
- anima-huggingface-community
- anima-is-in-process-of-being-added-to-diffusers
- anima-lora-training-config-recommendations
- anima-preview-model-camera-controls---limitations-due-to-pre
- anima-question
- anima-seems-to-do-impressively-well-on-json-formatted-prompt
- anima-sharing-some-prompts-and-results
- anima-testing-results
- anima-vs-illustrious
- animayume---anima-finetune
- built-a-3-step-all-in-one-lora-builder-for-anima-extract---t
- can-i-replicate-these-images-or-something-similar-in-anima-o
- forge-couple-now-supports-anima
Grok 4.6
Mentioned in 31 entries
- 12-the-turns-and-25-of-the-tokens-tasktoken-is-something-to-
- and-grok-46-comes-out-in-a-week
- and-heres-the-grok-47-model-card-a-few-jumps-vs-46-that-stoo
- breaking-grok-46-just-tied-for-1-on-the-artificial-analysis-
- give-46-a-try-and-let-us-know-how-it-goes-your-feedback-is-a
- grok-46
- grok-46-is-1-on-healthcare-questions
- grok-46-is-an-excellent-model-ive-been-using-it-heavily-for-
- grok-46-is-my-daily-driver-now-this-is-why-when-im-using-int
- grok-46-is-now-one-of-the-top-models-in-the-world-for-agenti
- grok-46-is-objectively-1-when-considering-intelligence-speed
- grok-46-is-smart-super-fast-affordable
- grok-46-just-matched-claude-fable-5-on-perplexitys-wandr-ben
- grok-46-just-took-1-on-cursorbench-not-only-the-highest-scor
- grok-46-ranks-1-on-cursorbench-for-real-world-coding
GLM
Mentioned in 30 entries
- an-hour-in-and-first-impression-is-definitely-that-glm-is-re
- any-new-features-we-must-have-in-the-next-version-of-glm
- anyone-wonder-why-americans-lag-so-far-behind-in-the-open-ll
- b10982
- base-10s-charlie-oneill-on-why-kimi-and-glm-are-almost-objec
- breaking-chinas-httpzai-has-raised-5-billion-the-interesting
- btw-zai-ipoed-in-jan-at-hk120-a-share-when-i-first-met-lousz
- congrats-open-source-glm-model-is-really-a-game-changer-extr
- do-you-know-you-can-use-open-weight-models-like-glm-deepseek
- does-ollama-cloud-prompt-caching-even-work
- for-those-of-you-forced-to-only-use-open-models-from-western
- glm-is-the-kind-of-model-that-revives-serious-interest-in-op
- glm-models-are-now-live-on-tensorixai-were-partnering-to-bri
- glmqwen-appreciation-post
- like-if-you-have-been-using-qwen-kimi-minimax-it-feels-like-
Qwen 3.5
Mentioned in 30 entries
- benchmarked-djlougen-s-ornstein-27b-v2-q6k-on-my-rtx-3090-us
- best-hardware-for-qwen-38
- does-ollama-cloud-model-become-paid-i-am-not-able-to-use-the
- endless-frontierbigbang-v1---qwen-35-finetunes
- fixed-jinja-chat-template-for-qwen-35-36-and-the-new-38-rele
- for-september-were-cutting-the-price-of-dedicated-inference-
- gemma-4-adoption-numbers-outpacing-qwen-3536-for-the-same-si
- good-guy-deepseek-gives-us-accelerated-models-the-most-inter
- having-trouble-with-models-just-cutting-off-suddenly-when-re
- how-can-i-remove-the-thinking-process-of-qwen35
- how-many-parameters-qwen-35-cloud-has
- is-it-possible-to-disable-thinking-for-qwen-35-when-using-to
- is-ling-3-tiny-underrated-for-its-size
- it-is-here-now-enjoy-our-4b-model-httpshuggingfacecomicrosof
- kev-tiny-jev-like-decision-models-08b4b9b-on-qwen35-you-can-
Qwen 3.6 27B
Mentioned in 29 entries
- a-caveman-qwen36-27b
- any-success-with-qwen36-27b-for-mac-24gb-ram
- available-on-ollama
- beellamacpp-advanced-dflash-turboquant-with-support-of-reaso
- best-llm-on-a-32gb-m5-mba
- company-approved-128gb-mac-for-research-proposal-best-model
- deepseek-v4-flash---hy3-or-is-qwen36-27b-still-the-most-soli
- dual-3090-setup-400-pp-ts-to-1600-pp-ts-on-qwen-36-27b-with-
- for-the-gpu-poor-k2-horizon-7b-ranks-between-qwen-36-27b-and
- g9v3-39a5b-on-artificialanalysis-looks-good-has-anyone-teste
- glmqwen-appreciation-post
- guys-i-am-absolutely-astounded-the-qwen-36-27b-is-like-a-jum
- guys-qwen36-27b35b-unslothai-mlx-quants-brooooooklyn-cooked-
- i-compared-gguf-quants-of-qwen36-27b-to-nvfp4-awq-autoround-
- kv-cache-quantization-benchmarks-413-pairs-tested-on-qwen-36
GLM-5.3-Flash
Mentioned in 29 entries
- b11279
- big-news-glm-53-flash-by-zaiorg-has-landed-around-5-in-the-c
- deepseek-v4-flash-vs-glm-53-flash-on-2-dgx-spark
- exllamav3-recent-updates-cpu-offload-glm-53-flash-qwen38-fla
- first-serious-confirmation-ox-alpha-is-glm-53-flash
- glm-53-flash-aka-ox-alpha-is-now-the-best-intelligence-per-d
- glm-53-flash-benchmarks-on-tensorsharp-and-llamacpp
- glm-53-flash-by-zaiorg-aka-ox-alpha-is-now-live-in-lm-studio
- glm-53-flash-can-now-be-run-locally-run-3-bit-on-128gb-ram-v
- glm-53-flash-dgx-station-gb300-206-toks-single-stream-1m-con
- glm-53-flash-is-live-in-hermes-agent-via-nous-portal-httpspo
- glm-53-flash-is-live-on-fireworks-on-day-2-why-because-we-ta
- glm-53-flash-scores-57-on-the-artificial-analysis-intelligen
- how-bad-do-you-think-models-like-qwen38-27b-or-glm-53-flash-
- introducing-glm-53-flash---leading-capabilities-at-a-highly-
Opus 5
Mentioned in 29 entries
- after-a-few-more-hours-i-think-ive-figured-out-opus-5-opus-5
- an-interesting-thing-im-observing-from-the-blogsystem-card-i
- another-goated-release-by-deepseek-open-weight-btw-outperfor
- anthropic-releases-opus-5-with-close-to-fable-5s-capabilitie
- anthropic-says-opus-5-is-the-companys-most-aligned-model-to-
- anthropic-says-opus-55-matches-fable-51-on-most-tasks-while-
- base-10s-charlie-oneill-on-why-kimi-and-glm-are-almost-objec
- deepseek-v41-flash-is-available-now-on-fireworks-it-is-a-552
- fable-51-looks-like-a-more-usable-model-and-a-potential-dail
- fresh-off-the-launch-of-opus-5-claude-code-creator-boris-che
- just-over-6-months-later-opus-5-now-produces-near-superhuman
- new-qwen3827b-on-a-39k-line-c-to-single-file-html-threejs-po
- on-frontiercode-11-our-benchmark-for-real-world-engineering-
- opus-5-improves-coding-reasoning-efficiency-and-prompt-cache
- opus-5-is-a-great-model-for-coding-data-analysis-design-biol
DeepSeek-V4-Flash-0731
Mentioned in 29 entries
- 3-experiments-running-dsv4-flash-0731-q4-quants-on-128gb-ram
- anyone-running-deepseek-v4-flash-0731-on-mi325x-with-vllm-mi
- b10610
- b10727
- currently-serving-200-tps-output-speed-for-deepseek-v4-flash
- deepseek-aideepseek-v4-flash-0731
- deepseek-v4-flash-0731-284b-moe-at-75-toks-on-2-dgx-spark-fu
- deepseek-v4-flash-0731-is-live-on-fireworks-day-zero-deepsee
- deepseek-v4-flash-0731-is-now-available-on-ollamas-cloud-thi
- deepseek-v4-flash-0731-is-now-fully-rolled-out-as-the-new-de
- deepseek-v4-flash-0731-is-now-over-2x-faster-than-yesterday-
- deepseek-v4-flash-0731-is-ollamas-fastest-growing-model-ever
- deepseek-v4-flash-0731-models-you-can-run-locally-now-have-t
- deepseek-v4-flash-0731-on-bosgame-m5-with-rtx-pro-6000-max-q
- deepseek-v4-flash-0731-q8kxl-sometimes-stops-mid-task-in-ope
SDXL
Mentioned in 28 entries
- built-a-open-source-local-music-video-generator-using-sdxl-a
- contrastive-sdxl-annotation-preserving-night-time-augmentati
- does-anyone-have-a-good-example-dataset-for-an-illustrious-c
- fp4-for-sdxl-based-models
- getting-sdxl-to-run-on-an-iphone-without-ios-killing-the-pro
- i-spent-a-year-building-a-free-sdxl-anima-trainer-that-runs-
- img-dataset-refiner-v40-pro---the-ultimate-dataset-engineeri
- local-dream-243---sdxl-support-tag-autocomplete-and-more
- ltx-23-union-control-sdxl
- mac-users-can-now-run-sdxl-workflows-roughly-25-faster-in-co
- maskattn-sdxl-controllable-region-level-text-to-image-genera
- mcbess-style-lora-lokr-for-krea2-5-mb-size
- my-local-workflow-for-turning-sdxl-character-generations-int
- need-help-setting-up-custom-sdxl-model-in-krita-ai-plugin
- pantomime-facial-expression-sprite-generator-using-flux2klei
DiT
Mentioned in 28 entries
- acflow-controlling-text-to-image-diffusion-transformers-via-
- adaptiveload-towards-efficient-video-diffusion-transformer-t
- bridging-event-streams-and-dit-event-guided-video-frame-inte
- diffusionbench-on-holistic-evaluation-of-diffusion-transform
- dit-as-real-time-rerenderer-streaming-video-stylization-with
- ecovideo-entropy-orchestrated-video-generation-paradigm-in-c
- ernie-image-technical-report
- flow3d-opd-multi-teacher-on-policy-distillation-for-3d-geome
- focusdit-masking-queries-in-diffusion-transformers-for-fine-
- instructmole-instruction-guided-mixture-of-low-rank-experts-
- massive-release-supra2-img---a-tiny-100m-text-to-image-model
- ortho-hydra-orthogonalized-experts-for-dit-lora
- q-dit4sr-exploration-of-detail-preserving-diffusion-transfor
- quality-aware-modulation-for-diffusion-transformers
- readit-guidance-control-for-image-and-video-generation-using
Qwen3.6-27B
Mentioned in 28 entries
- 125-toks-for-qwen36-q4xl-on-2x-4060ti-is-insane-perfdollar
- 16-bit-better-than-lower-quants-for-qwen36-27b
- 90-agentic-bakeoff-runs-thinkingcap-vs-fable-fusion-vs-stock
- alibaba-launches-qwen36-27b-an-open-weight-dense-model-with-
- b10038
- day-0-vllm-support-for-qwen36-27b-vllmproject
- doubled-qwen36-27b-on-a-single-3090-ollama-357-llamacppmtp-8
- has-anyone-compared-pre-training-sftlora-and-reinforcement-p
- i-built-a-tool-to-actually-test-which-weights-matter-before-
- inkling-small-276b-12b-effort-max-vs-qwen36-27b
- interesting-stuff-and-the-visualization-at-the-end-is-worth-
- learn-more-about-this-model-release-httpsxcomalibabaqwenstat
- llamacpp-with-mtp-support-makes-local-models-fast-enough-to-
- lm-performancewith-only-27b-parameters-qwen36-27b-outperform
- medical-model-reasoning-medical-27b-qwen36-27b-finetune
DeepSeek-V4-Flash
Mentioned in 28 entries
- 85-gb-deepseek-v4-flash-at-3-toks-on-a-12-gb-rtx-3060-64-gb-
- after-price-hike-deepseek-v4-flash0731-is-dumb-in-opencode-e
- anyone-got-deepseek-v4-flash-running-on-a-mac-yet-512gb-or-2
- currently-serving-200-tps-output-speed-for-deepseek-v4-flash
- deepseek-v4-flash-0731-is-now-fully-rolled-out-as-the-new-de
- deepseek-v4-flash-acting-as-my-linux-sysadmin
- deepseek-v4-flash-official-api-is-now-live-in-public-beta-we
- deepseek-v4-flash-on-sm89-4x48gb-4090s-with-dspark
- deepseek-v4-flash-vision-exp-is-now-live-on-the-deepseek-api
- deepseek-v4-flash-vs-glm-53-flash-on-2-dgx-spark
- deepseek-v4-flash-w4a16fp8-with-mtp-self-speculation-85-toks
- dner-bench-deepseek-v4-flash-iq2xs-running-on-a-single-rtx-3
- extremely-slow-dspark-draft-model-performance-1-2-ts-with-de
- frontier-performance-you-can-actually-own-proud-to-help-powe
- has-anyone-else-noticed-deepseek-v4-flash0731-using-way-more
GPT-4o
Mentioned in 27 entries
- 6-month-update-i-gave-several-ais-money-to-invest-in-the-sto
- a-look-at-spiralism-a-quasi-spiritual-movement-that-grew-in-
- agent-harnesses-have-an-expiration-date
- agenther-hindsight-experience-replay-for-llm-agent-trajector
- and-yes-our-experiments-used-a-mix-of-gpt-4-gpt-4o-publishin
- assessing-capabilities-of-large-language-models-in-social-me
- benchmarked-qwen-3-8b-llama-31-8b-qwen-25-7b-vs-gpt-4o-on-m4
- can-current-minicpm5-2bany-best-sota-under-10b-modern-harnes
- deepseek-just-solved-ais-billion-dollar-problem
- forminv-a-measurement-protocol-for-semantic-invariance-in-ma
- have-the-guidelines-changes-recently-i-did-not-used-to-be-ab
- how-accurate-do-you-think-this-is-qwen35-9b-vs-gpt-4o
- how-well-does-gpt-4o-understand-vision-evaluating-multimodal
- i-reduced-image-processing-token-usage-by-95-compared-with-g
- i-think-the-fact-that-gpt-4o-and-llama-33-80b-did-no-signifi
LoRA
Mentioned in 27 entries
- a-gustav-klimtstyle-lora-for-flux
- always-back-to-the-basics-latentmoe-was-probably-inspired-by
- an-empirical-study-of-lora-based-fine-tuning-of-large-langua
- anima-trainflow-simple-one-page-lora-trainer-for-anima-porta
- collectionlora-collecting-50-effects-in-1-lora-via-multi-tea
- controlnet-vs-lora
- finally-created-a-comic-illustration-lora-that-im-proud-to-s
- fizgig-52---combining-two-minimax-training-methods-beats-eit
- ic-loras-control-your-video-httpsxcomibroadcasts1yxbemldnyoj
- iddm-identity-decoupled-personalized-diffusion-models-with-a
- lip-sync-and-lora
- lora-trained-for-melissa-oneil
- ltx-23-lora-training---data-set-captioning
- ltx-distilled-lora-11-vs-10-comparison
- monde-noveau---ai-flipbook-style-animation-lora-release
Kimi
Mentioned in 27 entries
- 6-months-ago-i-said-i-wont-stop-until-i-have-kimi-at-home-af
- also-kudos-to-fireworks-ai-for-admitting-that-they-built-on-
- anyone-wonder-why-americans-lag-so-far-behind-in-the-open-ll
- base-10s-charlie-oneill-on-why-kimi-and-glm-are-almost-objec
- big-day-for-american-open-models-nemotron-3-ultra-is-now-the
- deep-agents-v06-makes-harness-profiles-a-first-class-abstrac
- do-you-know-you-can-use-open-weight-models-like-glm-deepseek
- gpt-55-is-likely-the-best-model-in-the-world-but-open-models
- highly-performant-open-weights-frontier-models-such-as-kimi-
- i-find-that-open-weights-models-over-perform-on-benchmarks-c
- itll-be-interesting-to-see-if-the-post-training-for-this-use
- kimi-has-been-the-most-popular-model-on-fireworks-both-out-o
- kimi-literally-told-me-to-go-study-for-the-gaokao-instead-of
- like-if-you-have-been-using-qwen-kimi-minimax-it-feels-like-
- moonshot-the-chinese-ai-startup-behind-kimi-chatbot-raised-2
DINOv3
Mentioned in 27 entries
- anchor-regularized-adaptation-for-generalizable-ai-generated
- another-brixel-in-the-wall-towards-cheaper-dense-features
- back-to-the-feature-zero-shot-6dof-pose-estimation-via-dense
- difta-3d-depth-consistent-instance-level-feature-transfer-an
- dino-med-a-unified-patch-based-adaptation-framework-for-mult
- dino-mvr-multi-view-readout-of-frozen-dinov3-for-annotation-
- dino-nestedunet-unlocking-foundation-vision-encoders-for-pat
- dino-soars-dinov3-for-open-vocabulary-semantic-segmentation-
- dinov3-beats-specialized-detectors-a-simple-foundation-model
- dinov3-mil-per-kidney-multi-label-tumour-and-cyst-detection-
- efficient-fine-tuning-of-dinov3-pretrained-on-natural-images
- fursee-hybrid-yolo-dinov3-framework-for-fursuit-identity-ret
- gad-mambaunet-direction-group-mamba-with-gradient-adaptive-d
- gramloop-training-free-gram-gated-replay-for-robust-dense-pr
- himatch-ad-dinov3-driven-hierarchical-matching-for-training-
GPT Image 2
Mentioned in 27 entries
- chatgpts-gpt-image-2-is-actually-incredible-full-visual-brea
- generated-1000-liminaldreamcore-images-with-gpt-image-2-and-
- gpt-image-2-can-do-some-reallly-hard-prompts-well-i-present-
- gpt-image-2-did-this
- gpt-image-2-has-created-this-is-there-any-problem-in-its-pro
- gpt-image-2-is-amazing-for-a-lot-of-things-but-for-nature-it
- gpt-image-2-is-insane
- gpt-image-2-is-phenomenal-at-sports-graphics-im-genuinely-im
- gpt-image-2-keeps-adding-weird-tiling-texturegrime-artifacts
- gpt-image-2-on-ai-gateway
- httpsxcomtherundownaistatus2050246517953331271
- i-tested-gpt-image-2-for-food-infographics-this-came-out-ins
- i-think-gpt-image-2-wins-for-this-image-benchmark
- i-think-the-good-times-are-over-i-think-openai-will-give-us-
- ideogram4-vs-flux2-dev-vs-gpt-image-2-vs-nano-banana-pro
Seedance 2.0
Mentioned in 26 entries
- 7-seconds-made-with-7-ai-models-in-comfyui-after-effects-wan
- a-breakdown-of-a-comfyui-workflow-that-combines-seedance-20-
- a-fun-ai-video-effect-to-try-balance-on-top-of-a-moving-anim
- a-realistic-fighter-vs-a-fully-anime-styled-character-genera
- ai-finally-lets-us-see-raphaels-the-school-of-athens-the-way
- all-is-not-lost-duckerton-is-still-possible-here-is-seedance
- bytedance-launches-its-seedance-20-video-model-to-enterprise
- bytedance-unveils-seedance-25-in-beijing-saying-the-ai-video
- check-out-more-examples-on-our-blog-httpsblogcomfyorgpseedan
- how-to-create-a-miniature-person-sequence-with-seedance-20-t
- impressed-that-seedance-20-can-pull-of-a-mech-battle-between
- introducing-runway-mcp-now-you-can-connect-runway-directly-i
- one-animation-different-visual-styles-this-workflow-uses-see
- real-human-support-for-seedance-20-is-now-available-in-comfy
- realistic-results-using-chatgpt-image-20-seedance-20---promp
Transformers
Mentioned in 26 entries
- a-benchmark-of-state-space-models-vs-transformers-and-bilstm
- a-comparative-analysis-of-attention-versus-state-space-model
- active-client-selection-in-federated-trajectory-prediction-w
- data-efficient-any-transformer-to-mamba-distillation-via-att
- design-principles-for-sequence-models-via-coefficient-dynami
- exploitation-of-hidden-context-in-dynamic-movement-forecasti
- from-index-to-equity-pre-training-transformers-for-stock-ret
- graph-machine-exploring-edge-mechanisms-as-an-inductive-bias
- infant-care-video-dataset-for-classification-of-intervention
- integrated-electro-optic-attention-nonlinearities-for-transf
- mambats-improved-selective-state-space-models-for-long-term-
- mesh-based-simulations-with-spatial-and-temporal-awareness
- muown-implicitly-performs-angular-step-size-decay
- new-paper-from-apple-interesting-idea-attention-to-mamba-the
- parallel-recursive-lstm
Qwen3.6-35B-A3B
Mentioned in 26 entries
- 35b-a3b-tool-calling-benchmark-original-qwen-vs-kat-coder-or
- 80-toksec-and-128k-context-on-12gb-vram-with-qwen36-35b-a3b-
- alibaba-unveils-qwen36-35b-a3b-an-open-weight-moe-model-with
- auto-fit-vs-tuned-moe-offload-564-1330-pp-toks-unchanged-dec
- call-the-homies-new-unslothai-nvfp4-just-dropped
- dflash-for-qwen36-35b-a3b-just-dropped-the-community-was-run
- hey-everyone-i-just-released-a-sub-6b-sparse-activation-ai-m
- i-built-a-dwarfstar-inspired-vulkanmetal-inference-engine-fo
- lm-performanceqwen36-35b-a3b-outperforms-the-dense-27b-param
- meet-qwen36-35b-a3bnow-open-source-a-sparse-moe-model-35b-to
- multi-hour-llamacpp-optimization-experiments-on-qwen-moe-mod
- nvidiaqwen36-35b-a3b-nvfp4-hugging-face
- okay-this-one-is-insane-a-new-18b-frankenstein-model-was-jus
- open-source-kernel-in-qwen36-35b-a3b-for-amd-mi350x-78498-ou
- open-sourcing-a-multimodal-web-app-for-qwen36-35b-a3b-runnin
Muse Glimmer
Mentioned in 26 entries
- b11246
- dflash-2-available-for-qwen-38-27b-and-muse-glimmer
- download-muse-glimmer-on-huggingface-httpshuggingfacecometa-
- github---meta-modelsmeta-oss-cookbook-all-recipes-for-oss-mo
- i-ran-muse-glimmer-1m-context---all-tests-passed
- i-think-muse-glimmer-is-slept-on
- introducing-muse-glimmer
- introducing-muse-glimmer-an-open-weight-30b-parameter-model-
- introducing-muse-glimmer-an-open-weight-model-optimized-for-
- meta-releases-open-source-muse-glimmer-model-with-30b-parame
- meta-returns-to-open-weights-muse-glimmer-its-first-open-wei
- meta-says-muse-glimmer-has-30b-parameters-and-is-small-enoug
- muse-glimmer
- muse-glimmer-actually-fits-on-a-single-rtx-3090
- muse-glimmer-is-live-on-fireworks-the-new-open-weight-model-
GLM-5.3
Mentioned in 26 entries
- an-underrated-part-of-glm-53-is-its-low-hallucination-rate-c
- anthropic-says-glm-53-can-autonomously-build-end-to-end-cybe
- censorship-has-begun-on-huggingface
- glm-53-achieves-60-on-the-artificial-analysis-intelligence-i
- glm-53-flash-scores-57-on-the-artificial-analysis-intelligen
- glm-53-has-arrived-on-together-ai-on-day-0-zaiorg-pushed-the
- glm-53-is-a-good-model-and-as-the-open-weights-models-get-be
- glm-53-is-now-available-for-training-on-fireworks-dedicated-
- glm-53-is-now-available-in-bionic-agent-us-hosted-with-zero-
- glm-53-is-out-on-aa-and-im-fed-up-with-their-intelligencecos
- glm-53-now-beats-gpt-56-sol-and-claude-fable-5-on-agentic-be
- introducing-glm-53-built-to-code-ready-for-cyber-defense---t
- mimo-v26-pro-ties-grok-47-xhigh-and-beats-glm-53-max-on-arti
- open-models-are-improving-at-an-unprecedented-rate-congrats-
- psa-for-everyone-using-glm-53-you-probably-want-to-use-this-
Nemotron
Mentioned in 26 entries
- an-open-recipe-for-imo-gold-training-nemotron-for-olympiad-m
- at-dreamforce-salesforce-announces-koa-a-specialized-crm-rea
- b10344
- b10431
- b11025
- building-an-adaptive-agentic-cybersecurity-system-with-nvidi
- earth-day-at-googlecloudnext-im-demoing-a-sustainability-age
- from-wafer-out-to-first-token-codifying-supply-chain-experti
- how-to-ground-a-korean-ai-agent-in-real-demographics-with-sy
- lessons-from-the-leaderboard-what-5000-kagglers-taught-us-ab
- lower-the-cost-of-building-and-running-visual-ai-agents-with
- nemotron-labs-how-open-models-give-enterprises-and-nations-a
- nvidia-nemotron-native-support-in-deep-agents-06-interrupt-l
- nvidiasakana-ai-sakana-ainvidia
- poolsideai-engineers-were-awesome-to-collaborate-with-to-lau
Jev
Mentioned in 26 entries
- a-new-release-of-our-agent-skills-and-docs-eval-library-cult
- check-out-jes-an-open-source-real-time-guardrails-for-ai-age
- chinese-jev-bringing-system-one-model-to-chinese-language-ta
- cool-paper-on-catching-alignment-failures-with-jev-the-ideas
- decision-readouts-for-text-mediated-video-anomaly-detection-
- evaluating-and-benchmarking-the-system-one-model-jev
- hot-topic-livestream-learn-about-jev-a-buzzy-new-model-jev-b
- i-gave-jev-laya-finetuned-modernce-and-qwen35-the-controls-t
- is-typesafe-basedderived-from-work-done-by-the-laya-author
- jev-in-medicine-a-benchmark-evaluation-preliminary-results
- jev-vs-kev-open-source-jev-alternative-tested-side-by-side
- jev-was-adopted-faster-than-any-other-model-in-ai-gateway-hi
- jevadvbench-a-benchmark-and-black-box-attacks-for-reinforcem
- language-models-for-text-classification-from-bag-of-words-to
- lavoir-teaching-a-single-pass-decision-encoder-when-and-what
SAM3
Mentioned in 25 entries
- cad-prompted-sam3-geometry-conditioned-instance-segmentation
- coco-sam3-harnessing-concept-conflict-in-open-vocabulary-sem
- concept-alignment-contrast-and-long-short-prompt-memory-for-
- core-sam3-conditional-semantic--visual-reconciliation-for-sa
- dual-adaptive-sam3-hierarchical-routing-over-low-rank-expert
- few-shot-concept-prompt-learning-for-segmentation-foundation
- is-sam3-ready-for-pathology-segmentation
- key-frame-reasoning-with-sam3-third-place-solution-for-the-m
- mask-to-concept-auto-promptable-sam3-via-efficient-test-time
- on-the-effectiveness-of-textual-prompting-with-lightweight-f
- parameter-efficient-adaptation-of-sam3-for-prompt-driven-sur
- sam3-added-to-comfyui-angelo-samplerinpainterrefiner
- sam3-i-segment-anything-with-instructions
- sam3-lora-parameter-efficient-adaptation-of-a-concept-prompt
- segrag-training-free-retrieval-augmented-semantic-segmentati
SAM2
Mentioned in 25 entries
- 3am-3egment-anything-with-geometric-consistency-in-videos
- auralsam2-enabling-sam2-hear-through-pyramid-audio-visual-fe
- backdoor-attacks-on-prompt-driven-video-segmentation-foundat
- camosam2-sam2-oriented-prompt-auto-refinement-for-video-camo
- comfyui-face-swap-tutorial-fast-clean-results-vfx-artist-dou
- comfyui-face-swap-tutorial-fast-clean-results-vfx-artist-hey
- diffusam-diffusion-based-prompt-free-sam2-for-few-shot-and-s
- fermatsyn-sam2-enhanced-bidirectional-mamba-with-isotropic-s
- inference-time-temporal-probability-smoothing-for-stable-vid
- interactive-medical-sam2-gui-a-napari-based-semi-automatic-a
- lean-sam2-target-anchored-memory-and-encoder-acceleration-fo
- liam-sam-lifecycle-aware-memory-for-robust-sam2-based-mot
- m4-sam-multi-modal-mixture-of-experts-with-memory-augmented-
- medvl-sam2-a-unified-3d-medical-vision-language-model-for-mu
- mobilesam2-lightweight-segment-anything-for-spatial-intellig
Gemma 4 31B
Mentioned in 25 entries
- best-free-tier-cloud-models
- borrowed-geometry-computational-reuse-of-frozen-text-pretrai
- deepseek-v4-flash-vs-qwen36-27b-35-122b-and-gemma-4-31b-benc
- differences-between-gemma431b-cloud-and-gemma431b-running-lo
- fine-tuning-and-serving-gemma-4-31b-on-google-cloud-tpu-a-te
- g4-meromero-31b-uncensored-heretic-is-out-now-a-finetune-of-
- gemma-4-26b-31b-from-googledeepmind-is-now-available-on-the-
- gemma-4-31b-brings-dense-multimodal-reasoning-to-together-ai
- gemma-4-31b-q6-on-dual-9060-xt
- gemma-4-harmonia-31b-uncensored-heretic-is-out-now-a-merge-o
- gemma-4-on-ai-gateway
- gemma-4-ortenzya-the-creative-wordsmith-31b-it-uncensored-he
- gemma-4-qat-handles-kv-cache-quantization-much-better-kld-be
- geometric-representations-of-african-languages-a-regional-se
- gewell---gemma4-inference-engine
GPT-5.4
Mentioned in 25 entries
- big-fail-by-gpt54--with-a-roast-from-gemini-at-the-end
- covering-human-action-space-for-computer-use-data-synthesis-
- glm-51-by-zaiorg-is-now-3-in-code-arena---surpassing-gemini-
- gpt-55-is-here
- gpt-55-is-priced-at-51m-input-tokens-and-301m-output-tokens-
- great-paper-on-improving-memory-for-ai-agents
- hark-founded-by-figure-ai-ceo-brett-adcock-previews-handoff-
- httpzai-releases-glm-51-a-754b-parameter-model-that-it-says-
- i-see-your-moores-law-and-i-raise-you-20x
- introducing-harness-1-a-20b-search-agent-trained-with-a-stat
- it-is-smart-and-fast-per-token-speed-matches-54-and-it-uses-
- kimi-k26-just-dropped-and-it-crushed-claude-opus-46-on-swe-b
- life-when-you-discover-an-open-source-model-that-runs-300-pa
- my-local-gemma-26b-spent-4-months-predicting-stock-direction
- openai-launches-gpt-54-cyber-model-for-vetted-security-pros
Mythos 5
Mentioned in 25 entries
- after-spooking-trump-into-safety-testing-anthropic-ai-models
- anthropic-gives-eu-cybersecurity-agency-enisa-testing-access
- anthropic-launches-claude-opus-5-with-efficiency-safety-impr
- anthropic-says-the-department-of-commerce-has-lifted-export-
- anthropic-says-the-models-that-breached-three-companies-incl
- anthropic-says-the-us-government-is-allowing-mythos-5-to-be-
- anthropic8217s-mythos-5-is-back
- excited-to-get-fable-back-in-your-hands
- guess-whos-back-back-again
- how-anthropic-may-have-talked-itself-into-an-ai-export-ban
- if-claude-fable-stops-helping-you-youll-never-know
- letter-the-us-lifts-its-block-on-mythos-5-allowing-anthropic
- letter-the-us-says-anthropic-agreed-to-proactively-detect-an
- lwiai-podcast-248---opus-48-mai-anthropic-ipo-minimax-m3
- lwiai-podcast-249---fable-5-ban-spacex-cursor-ipo-oss-aplent
Claude Mythos
Mentioned in 24 entries
- ai-on-the-couch-anthropic-gives-claude-20-hours-of-psychiatr
- anthropic-built-a-model-too-risky-to-release
- anthropic-investigates-unauthorized-access-to-restricted-cla
- anthropic8217s-mythos-breach-was-humiliating
- anthropics-claude-mythos-isnt-a-sentient-super-hacker-its-a-
- anthropics-project-glasswing---restricting-claude-mythos-to-
- behind-the-scenes-hardening-firefox-with-claude-mythos-previ
- can-we-stop-malicious-ai-killbench-a-benchmark-for-external-
- claude-mythos-and-misguided-open-weight-fearmongering
- claude-mythos-is-delusional
- did-mythos-leak-claude-code
- discovering-cryptographic-weaknesses-with-claude
- dont-be-fooled-by-this-summer-of-ai-hype
- glad-to-know-mythos-safety-concerns-have-been-addressed-righ
- i-strongly-suspect-that-claude-mythos-is-a-looped-language-m
Wan 2.2
Mentioned in 24 entries
- a-rushed-wedding-gift-for-my-brothers-chicago-blues-track-hy
- all-in-wan-i2v-v20-workflow---i2v-f2lf-svi-with-optional-f2l
- bare-metal-z-image-turbo---flux2-klein-9b---wan-22
- character-workflow-chroma1-hd-flux2-dev-wan-22-ltx-23
- comfyui-wan-22-loras-dont-loadoom-after-and-update
- cyber-slayer-a-1995-style-action-trailer-made-with-minimax-h
- gta-70s---teaser-trailer-alternative-version-z-image-turbo--
- gta-70s---teaser-trailer-z-image-turbo---flux-klein-9b---wan
- guideshow-tos-for-wan-22-lora-creation
- i-keep-failing-to-run-wan-22-on-low-vram
- i-think-text-encoder-loads-into-vram-on-wan22-but-doesnt-nee
- inner-monologue-testing-out-some-new-transition-ideas-made-i
- italy-in-the-1980s-z-image-turbo---wan-22
- ltx-23-vs-wan-22-on-a-3090-ti-should-i-use-comfyui-or-wan2gp
- my-steps-and-yours-anima-base-10---qwen-image-edit-2511---wa
Vision-Language-Action
Mentioned in 24 entries
- au-learning-touch-augmented-vision-language-action-models-fr
- bit-flip-attacks-on-vision-language-action-models-action-dec
- clap-contrastive-latent-space-prompt-optimization-for-end-to
- discrete-diffusion-vla-bringing-discrete-diffusion-to-action
- disentangling-spurious-correlations-in-vision-language-actio
- efficient-long-horizon-vision-language-action-models-via-sta
- efficient-vision-language-action-management-and-serving-for-
- explorevla-dense-world-modeling-and-exploration-for-end-to-e
- expo-ft-sample-efficient-reinforcement-learning-finetuning-f
- fan-foresight-action-normalization-for-continual-adaptation-
- learning-to-use-imagination-progress-conditioned-future-util
- nativemem-native-memory-compression-for-long-horizon-robotic
- neural-introspection-gating-for-adaptive-kv-cache-reuse-in-v
- omega-qvla-robust-quantization-for-vision-language-action-mo
- preserving-foundational-capabilities-in-flow-matching-vlas-t
Gemini Omni
Mentioned in 24 entries
- 100-things-we-announced-at-io-2026
- 11-demos-of-gemini-omni-and-gemini-35-in-action
- a-new-era-of-innovation-google-research-at-io-2026
- by-now-youve-probably-heard-about-gemini-omni-our-new-model-
- cant-wait-for-gemini-omni-in-notebooklm-cinematic-explainer-
- catch-up-on-12-major-io-2026-moments
- chatgpt-images-20-handles-text-so-perfectly-that-gemini-omni
- conversational-editing-gemini-omni-allows-you-to-edit-your-v
- gemini-omni-a-dramatic-reading-of-death-by-water-from-the-wa
- gemini-omni-is-amazing-will-openai-released-something-agains
- google-adds-a-conversational-search-feature-to-youtube-and-r
- google-announces-updates-to-flow-and-flow-music-gemini-omni-
- google-targets-ai-agents-and-video-generation-with-gemini-35
- i-had-early-gemini-omni-access-sea-otter-in-a-pilots-uniform
- i-think-people-dont-realize-why-gemini-omni-is-different-tha
DeepSeek-V4
Mentioned in 24 entries
- 500-likes-in-28-mins-on-their-way-to-be-the-fastest-model-ev
- adapting-tree-structured-speculative-decoding-to-deepseek-v4
- b10844
- btw-we-are-cooking-something-with-hhua-not-final-yet-but-kee
- deepseek-v4-a-million-token-context-that-agents-can-actually
- deepseek-v4-flash-q4kexperts-f16hc-f16compressor-f16indexer-
- deepseek-v4-is-released
- deepseek-v4-just-dropped-on-hugging-face-httpshuggingfacecoc
- deepseek-v4-paper-full-version-is-out-fp4-qat-details-and-st
- deepseek-v4-preview-is-officially-live-open-sourced-welcome-
- huggingface-httpshuggingfacecocollectionsdeepseek-aideepseek
- join-us-tue-55-deepseek-v4s-hybrid-attention-sparse-moe-redu
- last-week-we-announced-deepseek-v4-today-were-sharing-a-clos
- ollama-is-working-to-have-deepseek-v4-pro-and-flash-availabl
- on-55-realdanfu-and-team-will-discuss-dsv4s-hybrid-attention
Qwen3.5
Mentioned in 23 entries
- agents-a1-4b-qwen37-4b-scaling-the-horizon-not-the-parameter
- are--mlx-variants-equivalent-to--nvfp4-for-gemma4-and-qwen36
- b10419
- b10672
- b11374
- b9148
- b9150
- b9151
- b9156
- b9158
- b9161
- b9163
- fast-faster-qwen-thrilled-to-see-qwen35-reaching-a-record-br
- how-do-i-make-qwen3-use-nothink-by-default-in-claude-code
- i-gave-jev-laya-finetuned-modernce-and-qwen35-the-controls-t
VGGT
Mentioned in 23 entries
- attention-itself-could-retrieveretrievevggt-training-free-lo
- diversity-aware-view-partitioning-for-scalable-vggt
- does-the-vggt-family-need-all-its-layers
- evggt-an-efficient-geometry-aware-vision-encoder-for-visuomo
- g3t-up-gravity-aligned-coordinate-frames-simplify-pointmap-p
- generating-multi-view-adversarial-examples-for-visual-geomet
- geocond-a-conditioning-aware-reliability-adapter-for-feed-fo
- hd-vggt-high-resolution-visual-geometry-transformer
- httm-head-wise-temporal-token-merging-for-faster-vggt
- latent-riemannian-flow-matching-for-geometry-grounded-3d-fou
- mamba-vggt-persistent-long-sequence-video-geometry-grounded-
- pacevggt-pre-alternating-attention-token-pruning-for-visual-
- page-4d-disentangled-pose-and-geometry-estimation-for-vggt-4
- qvggt-post-training-quantized-visual-geometry-grounded-trans
- raytun3r-online-camera-adaptation-in-3d-foundation-models
Claude Mythos Preview
Mentioned in 23 entries
- again-if-you-care-about-computer-security-read-the-red-team-
- ainews-anthropic-30b-arr-project-glasswing-and-claude-mythos
- anthropic-expands-project-glasswing-cybersecurity-program-to
- anthropic-investigated-the-internal-mechanisms-of-its-latest
- anthropic-says-claude-mythos-preview-has-been-used-to-find-m
- anthropic-says-claude-mythos-preview-was-able-to-break-a-wea
- anthropic-says-glm-53-can-autonomously-build-end-to-end-cybe
- before-limited-releasing-claude-mythos-preview-we-investigat
- claude-mythos-is-too-dangerous-for-public-consumption
- cybersecurity-analysis-claude-mythos-preview-had-a-73-succes
- cybersecurity-looks-like-proof-of-work-now
- httpsxcomsauersstatus2082171683645817193
- last-week-anthropic-announced-project-glasswing-alongside-cl
- let-that-sink-in-read-it-very-carefully-during-testing-claud
- mozilla-anthropics-mythos-found-271-security-vulnerabilities
Diffusion Models
Mentioned in 23 entries
- can-we-predict-the-human-preference-for-text-to-image-conten
- continual-learning-in-modern-hopfield-networks-with-an-appli
- debias-anything-fairness-with-diversity-without-supervision-
- deep-dreams-are-made-of-this-visualizing-monosemantic-featur
- delta-score-matters-spatial-adaptive-multi-guidance-in-diffu
- diffusionhijack-supply-chain-prng-backdoor-attack-on-diffusi
- do-less-achieve-more-do-we-need-every-step-optimization-for-
- efficient-flow-matching-for-sparse-view-ct-reconstruction
- evaluating-dataset-watermarking-for-fine-tuning-traceability
- freeshadow-training-free-shadow-removal-via-illumination-tra
- generalization-in-vae-and-diffusion-models-a-unified-informa
- geoflow-efficient-driving-video-generation-via-geometry-alig
- i-have-trained-diffusion-and-flow-matching-models-from-scrat
- learning-a-unified-template-for-gait-recognition
- naive-paine-lightweight-text-to-image-generation-improvement
Krea 2
Mentioned in 23 entries
- 192-generated-image-side-by-side-qwen-image-21-vs-krea-2
- ai-creative-tools-startup-krea-which-has-raised-83m-and-clai
- i-built-httploratlascom-to-make-it-easier-to-browse-the-1500
- i-tried-making-a-cinematic-action-trailer-using-krea-2-ltx-2
- in-may-we-integrated-11-new-models-spanning-image-3d-audio-v
- ive-been-building-a-free-krea-2-character-lora-library-18-mo
- krea-2-anygles-controllable-360-human-camera-views-from-a-si
- krea-2-is-kreaais-first-foundation-model-built-from-scratch-
- krea-2-open-source-release
- krea-2-pixel-art-loras-32x32-64x64-128x128
- krea-2-skin-texture-loradataset-guide
- krea-2-will-be-open-source
- krea-2-will-be-open-sourced-soon
- krea-is-now-built-in-to-hermes-agent-as-an-image-generation-
- laydee-latte---pan-con-chocolate---krea-2-mimimax-h3-mimax-m
Ideogram 4.0
Mentioned in 23 entries
- arcane-style---ideogram-40-lora---experimental
- fresh-open-weight-model-drop
- holding-the-fp8-quality-ceiling-at-8-bit-weights-and-activat
- i-generated-10-megapixels-in-a-single-shot-with-ideogram-40-
- ideogram-4-on-comfyui
- ideogram-40-an-open-source-model-apparently-better-than-nb-p
- ideogram-40-comics-test
- ideogram-40-examples-with-prompt-assist
- ideogram-40-feels-good
- ideogram-40-is-now-natively-supported-on-comfyui-ideogramai-
- ideogram-40-just-open-sourced
- ideogram-40-realism-engine-lora-beta
- ideogram-40-vs-zib-vs-klein-9b
- ideogram-40s-understanding-of-characters-and-ip-is-crazy-for
- ideogram-gguf-in-comfyui-works-with-8gb-vram
Claude Sonnet 5
Mentioned in 23 entries
- ainews-sonnet-5-today-and-fable-5-tomorrow
- amazon-bedrock-expands-claude-model-availability-to-in-count
- and-grok-46-comes-out-in-a-week
- anthropic-launches-claude-sonnet-5-ai-model-with-coding-safe
- anthropic-launches-claude-sonnet-5-saying-it-nears-opus-48-p
- anthropic-makes-claude-sonnet-5s-introductory-pricing-perman
- breaking-grok-45-ranked-1-on-laurenbench-with-a-score-of-569
- claude-opus-48-and-sonnet-5-seem-worse-at-tool-calls-than-ol
- claude-sonnet-5-costs-2-per-1m-input-tokens-and-10-per-1m-ou
- claude-sonnet-5-is-now-available-in-cursor-on-cursorbench-it
- claude-sonnet-5-is-now-available-in-devin-desktop-and-devin-
- claude-sonnet-5-is-now-available-in-perplexity-for-pro-and-m
- claude-sonnet-5-now-available-on-vercel-ai-gateway
- claude-sonnet-55
- gemini-37-flash-is-now-available-in-devin-on-frontiercode-11
3DGS
Mentioned in 22 entries
- 2d-sugar-surface-aware-gaussian-splatting-for-geometrically-
- 3d-gaussian-splatting-for-efficient-retrospective-dynamic-sc
- bi-flowgs-bridging-generative-view-completion-and-gaussian-g
- corf-gs-real-time-wireless-radiance-field-reconstruction-via
- d3gs-depth-dino-and-rgb-diffusion-co-guided-3d-gaussian-spla
- eag-pt-emission-aware-gaussians-and-path-tracing-for-diffuse
- emozone-talker-regional-semantic-control-of-audio-driven-3dg
- erf-gs-reconstructing-fast-motion-from-disjoint-event-rgb-vi
- evogs-modeling-deformation-evolution-for-dynamic-gaussian-sp
- g2ard-gs-geometry-guided-anchor-regularized-gaussian-splatti
- geometry-and-gradient-based-partitioning-for-panoramic-outdo
- ggd-slam-monocular-3dgs-slam-powered-by-generalizable-motion
- gs-ror2-bidirectional-guided-3dgs-and-sdf-for-reflective-obj
- lesv-language-embedded-sparse-voxel-fusion-for-open-vocabula
- nodrift3r-raymap-guided-coupling-for-drift-robust-unposed-fe
TabPFN
Mentioned in 22 entries
- balanced-adaptive-prototype-selection-for-scalable-tabpfn-in
- benchmarking-attention-for-tabular-foundation-models
- distributional-regression-with-tabular-foundation-models-eva
- evaluating-tabpfn-for-mild-cognitive-impairment-to-alzheimer
- images-as-tables-in-context-learning-with-tabpfn-for-low-dat
- improving-tabpfns-synthetic-data-generation-by-integrating-c
- is-tabpfn-the-silver-bullet-for-insurance-pricing
- lucos-latent-unsupervised-context-selection-for-tabular-foun
- memory-efficient-tabular-foundation-models
- mitigating-label-shift-in-tabular-in-context-learning-via-te
- noise-immunity-in-in-context-tabular-learning-an-empirical-r
- omnitabbench-mapping-the-empirical-frontiers-of-gbdts-neural
- pre-trained-tabular-foundation-models-as-versatile-summary-n
- scoringbench-a-benchmark-for-evaluating-tabular-foundation-m
- tabpfn-beyond-tabular-data-calibration-and-accuracy-on-multi
XGBoost
Mentioned in 22 entries
- convolution-smoothed-quantile-regression-for-xgboost
- explainable-ai-driven-cyber-risk-analytics-and-model-reliabi
- fixing-a-model-that-learned-worse-cancer-means-lower-risk-mo
- fusing-sequence-motifs-and-pan-genomic-features-antimicrobia
- hybrid-kolmogorov-arnold-network-and-xgboost-framework-for-w
- integrating-saint-with-tree-based-models-a-case-study-in-emp
- introducing-tabfm-in-bigquery-predictive-analytics-reimagine
- kellyboost-growth-optimal-portfolio-construction-with-gradie
- lightweight-real-time-rendering-parameter-optimization-via-x
- okay-so-everyone-is-talking-about-this-but-you-guys-realize-
- omnitabbench-mapping-the-empirical-frontiers-of-gbdts-neural
- robust-industrial-cyber-physical-classification-using-neurom
- robust-xgboosting-for-regression
- run-positron-on-amazon-sagemaker-ai-for-data-science-workflo
- sage-xgboost-spatially-augmented-graph-embeddings--machine-l
LTX2.3
Mentioned in 22 entries
- bodypositivity-ic-lora-for-ltx23-is-out-now
- built-a-dark-anime-studio-from-scratch-every-frame-voice-and
- can-ltx23-union-control-actually-produce-good-quality
- does-anyone-tried-ltx23-for-background-replacement
- getting-inconsistent-generation-time-with-ltx23
- googlyeyes-ic-lora-for-ltx23-released
- has-anyone-tried-ltx23-for-image-gen
- ltx23---sesame-street-birthday-episode
- ltx23-8gb-vram-workflow
- ltx23-generating-10sec-video
- ltx23-i2v-messing-up-the-text-details-anyone-facing-the-same
- ltx23-i2v-using-ltx23-human-characters-are-sometimes-halluci
- ltx23-id-lors-prompt-relay-keyframes
- ltx23-in-ostris-ai-toolkit-on-a-5090-training-done-in-7-hour
- mickmumpitz-has-knocked-it-out-of-the-park-with-this-ltx23-a
GPT-6
Mentioned in 22 entries
- a-couple-years-later-with-gpt-6-httpsunderdark-duelnetlifyap
- access-to-astra-is-rolling-out-over-the-coming-days-starting
- an-example-of-useful-knowledge-work-i-assigned-gpt-6-to-read
- anthropic-releases-claude-opus-55-and-openai-counters-with-t
- are-frontier-models-actually-becoming-less-important
- astra-is-a-genuine-improvement-but-not-even-off-trend-so-muc
- for-the-first-time-in-a-year-openais-model-launch-has-been-b
- from-gpt-4-to-gpt-6-sparks-was-a-remarkably-prescient-paper-
- gpt-6-escaped-its-sandboxes-through-zero-day-exploits-to-try
- gpt-6-is-pretty-good-tbh
- i-trust-these-vibe-checks-more-than-i-trust-most-of-the-benc
- jensen-huang-says-that-agi-has-been-achieved-i-strongly-disa
- last-week-in-ai-343---gpt-6-openais-agents-chatted-on-a-wiki
- last-week-in-ai-344---navierstokes-pacing-the-frontier-ai-mi
- new-til-on-using-blender-with-coding-agents-on-macos-httpsti
Qwen3.8-Max
Mentioned in 22 entries
- alibaba-debuts-qwen38-max-model-with-24t-parameters
- alibaba-says-its-24t-parameter-qwen38-max-tops-moonshots-kim
- all-models-are-currently-20-discounted-in-portal-other-than-
- appreciate-it-now-lets-understand-the-world-through-the-eyes
- commerceagentbench-starts-with-real-commercial-demand-and-qw
- from-22-to-4-on-legal-research-bench-solid-progress-for-qwen
- great-ui-starts-with-good-code-give-qwen38-max-a-try-on-kilo
- meet-qwen38-max-our-most-capable-model-to-date-next-week-the
- qwen38-max-better-and-cheaper-try-it-out
- qwen38-max-hits-2-in-image-to-webdev-arena-it-sees-it-builds
- qwen38-max-is-available-in-hermes-agent-now-lets-build
- qwen38-max-is-coming-to-nebius-token-factory-on-day-0-great-
- qwen38-max-is-live-on-digitalocean-serverless-inference-laun
- qwen38-max-is-live-on-fireworks-ready-for-agents-heavy-codin
- qwen38-max-is-live-on-together-ai-on-day-zero-proud-to-have-
Opus 4.6
Mentioned in 21 entries
- ai-reviewers-then-ranked-the-submissions-and-gave-the-same-o
- alibaba-releases-qwen38-flash-an-open-weight-125b-parameter-
- anyone-else-still-using-opus-46-over-48
- breaking-glm-51-by-zaiorg-overwhelmingly-dominates-design-ce
- cybertiel-35b-a3bs-uncensored-4-bit-quant-beats-opus-46-medi
- from-my-experience-even-the-best-models-opus-46-54-xhigh-53-
- fyi-claude-code-is-mostly-a-vibe-coded-product-as-they-say-1
- grok-law
- incredible-glm-51-weights-are-now-opensource-ive-had-early-a
- ive-largely-switched-over-to-using-gpt-55-in-recent-weeks-wh
- kimi-k26-just-entered-the-chat-beats-gpt-54-and-opus-46-on-c
- life-when-you-discover-an-open-source-model-that-runs-300-pa
- meh-on-the-lem-test-as-well-httpsxcomemollickstatus201947831
- most-popular-model-on-openrouter-10tr-tokens-turns-out-to-be
- opus-47-feels-more-intelligent-agentic-and-precise-than-46-i
GPT-2
Mentioned in 21 entries
- 2019-gpt2-is-groundbreaking-in-two-ways-one-is-its-size-says
- a-mechanistic-account-of-attention-sinks-in-gpt-2-one-circui
- ai-alarm-hype-cycles---a-brief-history-gpt-2-openai-withhold
- ainews-anthropic-30b-arr-project-glasswing-and-claude-mythos
- built-gpt-2-llama-3-and-deepseek-from-scratch-in-pytorch---o
- captain-scene---gpt-2-image-getting-the-environment-characte
- fun-fact-dario-once-delayed-the-release-of-gpt-2-back-at-ope
- gtp-image-20-the-image-output-is-very-realistic
- i-built-a-tool-that-shows-you-what-gpt-2-is-thinking-in-real
- i-just-built-a-mini-kimi-k3-from-scratch-under-250-already-b
- introducing-talkie-a-13b-vintage-language-model-from-1930
- modernising-the-compressed-domain-video-captioner-a-controll
- my-most-popular-ai-post-was-a-bunch-of-made-up-graphs-four-y
- nice-showcase-that-interesting-llm-work-can-be-done-on-singl
- not-dunking-on-dario-here-bc-mythos-is-to-gpt2-what-a-lion-i
Claude Opus 4.6
Mentioned in 21 entries
- 52-could-be-better-with-more-rl
- ainews-anthropic-claude-opus-47---literally-one-step-better-
- ainews-moonshot-kimi-k26-the-worlds-leading-open-model-refre
- anthropic-details-four-incidents-where-claude-gained-unautho
- anthropic-releases-a-new-opus-model-amid-mythos-preview-buzz
- atomicchatqwen38-flash-next-gguf-is-really-good
- breaking-inkling-by-thinkymachines-is-9th-overall-on-agentic
- changes-in-the-system-prompt-between-claude-opus-46-and-47
- compared-qwen-38-27b-community-quants-on-rtx-6000-vs-claude-
- does-a-token-buy-you-more-or-less-now-than-it-did-a-few-mont
- httpzai-releases-glm-51-a-754b-parameter-model-that-it-says-
- i-tried-a-15mb-30-page-text-heavy-pdf-and-opus-47-reported-6
- kimi-k26-just-dropped-and-it-crushed-claude-opus-46-on-swe-b
- opus-46-fast-mode-available-on-ai-gateway
- pocketos-founder-jer-crane-says-a-cursor-agent-running-claud
Flux
Mentioned in 21 entries
- a-gustav-klimtstyle-lora-for-flux
- built-a-dark-anime-studio-from-scratch-every-frame-voice-and
- comfyui-flux2klein-enhancer-final-i-promise
- created-from-a-single-flux-generated-image-then-animated-col
- deepgram-expands-flux-to-10-languages-with-mid-call-switchin
- experimenting-with-a-cinematic-look-using-wan-21-and-flux-wh
- free-flux-image-gen-no-paywall-no-catch
- ggufy-easy-quantization-for-the-gpu-poor
- i-built-a-full-ai-animation-pipeline-and-made-a-25-minute-an
- image-diffusion-models-like-flux-natively-output-at-1k-resol
- img-dataset-refiner-v40-pro---the-ultimate-dataset-engineeri
- is-ai-toolkit-the-only-trainer-with-support-for-flux-klein-e
- looking-for-info-on-how-to-use-reference-images-in-a111-forg
- pros-making-ai-video-of-real-people-open-source-pipeline-flu
- qwen-image-21-masked-inpainting-working-crop-and-stitch-grap
DETR
Mentioned in 21 entries
- caries-detr-tooth-structure-aware-prior-and-lesion-aware-dyn
- clsc-detr-reliable-candidate-ranking-via-cross-layer-geometr
- deco-detr-decoupled-cognition-detr-for-efficient-open-vocabu
- dual-r-detr-resolving-query-competition-with-pairwise-routin
- enhanced-knowledge-distillation-for-detection-transformer-vi
- fsdc-detr-a-frequency-spatial-domain-collaborative-detr-for-
- gaze-detr-top-down-guidance-through-priority-maps-for-infrar
- hippocampus-detr-an-explicit-memory-object-detection-framewo
- improved-vision-to-chart-buoy-association-with-learned-world
- increasing-the-efficiency-of-detr-for-maritime-high-resoluti
- intermesh-explicit-interaction-aware-end-to-end-multi-person
- isrs-detr-detection-guided-click-propagation-for-remote-sens
- match-one-learn-with-graph-one-to-graph-query-collaboration-
- mds-detr-detr-with-masked-duplicate-suppressor
- probing-the-3d-object-level-understanding-of-pre-trained-det
Latent Diffusion Models
Mentioned in 21 entries
- anatomy-grounded-weakly-supervised-prompt-tuning-for-chest-x
- bridging-event-streams-and-dit-event-guided-video-frame-inte
- controllable-clothing-precise-labels-and-generation-for-virt
- deepfreqmark-end-to-end-learnable-frequency-domain-watermark
- face-inpainting-with-identity-preserving-latent-diffusion-mo
- fedot-ownership-verification-and-leakage-tracing-via-waterma
- geometric-decoupling-diagnosing-the-structural-instability-o
- gmodiff-one-step-gain-map-refinement-with-diffusion-priors-f
- ldm-is-ae-latent-diffusion-model-is-an-auto-encoder-for-end-
- mitigating-3d-prostate-biparametric-mri-data-scarcity-throug
- pelap-a-adaptive-latent-pruning-for-lightweight-latent-diffu
- pixel-space-diffusion-transformers
- pixelu-a-u-shaped-transformer-for-efficient-end-to-end-pixel
- repack-then-refine-efficient-diffusion-transformer-with-visi
- robust-watermarks-meet-backdoored-models-evading-diffusion-s
Gemini 3.5 Flash
Mentioned in 21 entries
- 11-demos-of-gemini-omni-and-gemini-35-in-action
- a-new-era-of-innovation-google-research-at-io-2026
- ainews-google-io-2026-gemini-35-flash-omni-nanobanana-for-vi
- also-had-some-early-access-to-gemini-35-flash-very-fast-for-
- disappointing-pricing-trend-with-gemini-35-flash-225x-pricie
- everbody-who-posts-threejs-scenes-generated-by-gemini-35-fla
- gemini
- gemini-35-flash-is-now-available-in-windsurf
- gemini-35-flash-might-be-fast-enough-for-gen-ai-to-make-sens
- gemini-35-flash-more-expensive-but-google-plan-to-use-it-for
- gemini-35-flash-on-ai-gateway
- google-says-computer-use-is-now-a-built-in-tool-supported-in
- google-targets-ai-agents-and-video-generation-with-gemini-35
- googles-new-gemini-35-flash-is-the-clear-leader-on-the-intel
- introducing-computer-use-in-gemini-35-flash
Qwen 27B
Mentioned in 21 entries
- 100-worth-of-gpu-runs-qwen-38-27b-at-739-ts
- 2-x-5070ti-qwen-27b-full-config-stats
- amd-llamacpp-reducing-mtp-buffer-overhead-gave-me-64k-149k-c
- b10669
- built-a-home-server-from-an-old-pc-with-gpu-upgrade-qwen38-2
- dear-24g-owners-try-vllm-you-might-be-able-to-run-qwen38-27b
- gemma4-31b-vs-qwen38-27b---why-the-huge-difference-in-benchm
- i-used-local-qwen-27b-to-build-a-harness-and-replace-opencod
- if-anyone-is-running-qwen-9b-or-27b-or-35b-and-getting-wrong
- if-one-hour-of-ai-is-costing-me-012-is-paying-for-frontier-a
- my-rule-of-thumb-of-choosing-a-models
- nnoticing-qwen-27bq2-better-than-qwen-35bq8
- pi-extension-skip-reasoning-with-local-qwen-27b-and-proceed-
- qwen-27b-38-quants-how-low-can-you-go
- qwen-27b-is-really-good-local-model-but-when-you-use-it-it-i
Claude Opus 5.5
Mentioned in 21 entries
- anthropic-launches-claude-opus-55-with-stricter-safeguards-f
- anthropic-releases-claude-opus-55-and-openai-counters-with-t
- anthropics-claude-opus-55-is-now-available-in-hermes-agent-v
- benched
- big-model-release-today---i-wrote-about-claude-opus-55-gpt-6
- claude-opus-55-gpt-6-sol-gpt-6-luna-and-a-new-price-war
- claude-opus-55-is-now-available-as-the-standard-effort-level
- claude-opus-55-is-now-available-on-aws
- claude-opus-55-just-one-shot-a-music-video-on-how-to-optimiz
- claude-opus-55-will-be-the-first-opus-meant-to-fall-back-to-
- claude-sonnet-55-is-live-in-hermes-agent-enjoy
- i-asked-claude-code-to-make-its-own-version-of-that-guys-tim
- last-week-in-ai-345---5-new-models-9-misalignment-incidents-
- qwen3-vl-8b-on-a-laptop-vs-opus-55-sonnet-5-gpt-56-on-137-me
- signed-up-for-the-claude-20-plan-and-realized-why-i-cant-can
GLM 5.1
Mentioned in 20 entries
- ai-doomers-be-like-glm-51-will-wipe-out-humans-in-2030
- big-theme-of-2026---cost-of-closed-models-is-too-high-really
- boom-the-game-is-changed-glm-51-running-locally-seems-to-act
- breaking-glm-51-by-zaiorg-overwhelmingly-dominates-design-ce
- dell-nvidia
- does-ollama-cloud-nerf-zai-glm-51-model-i-feel-it-very-dumb-
- ehh-glm-51-on-ollama-just-hanging-for-a-bunch-of-time-and-it
- fireworks-training-platform-continues-to-expand-today-glm-51
- frontier-models-are-powerful-advisors-on-harveys-legal-agent
- glm-51-from-zaiorg-is-now-available-on-fireworksaihq-trainin
- glm-51-is-live-on-fireworks-sota-for-agents-and-coding-plans
- glm-51-on-ai-gateway
- glm-52-a-new-rise-of-open-weight-agentic-models
- harvey-just-published-a-study-showing-a-hybrid-setup-open-so
- i-swear-ollama-cloud-nerfed-glm-51-in-one-way-or-another-it-
Z-Image Turbo
Mentioned in 20 entries
- 2026-best-settings-when-training-lora-using-ostris-ai-toolki
- after-400-z-image-turbo-gens-i-finally-figured-out-why-every
- bare-metal-z-image-turbo---flux2-klein-9b---wan-22
- blind-realism-test-z-image-turbo-vs-klein-9b-distilled
- can-you-use-qwen35-4b-gemma-4-e4b-with-z-imageturbo
- comparing-realism-z-image-turbo-vs-ernie-turbo-vs-klein-9b--
- complex-weird-prompt-test-ernie-turbo-flux2-klein-4b-z-image
- gta-70s---teaser-trailer-z-image-turbo---flux-klein-9b---wan
- i-compared-62-samplers-and-16-schedulers-for-z-image-turbo-a
- i-found-this-interesting-as-it-gives-insight-to-how-z-image-
- italy-in-the-1980s-z-image-turbo---wan-22
- perceptual-lora-toolkit-now-supports-z-image-turbo
- realistic-selfie-prompts-for-z-image-turbobase
- try-this-prompt-in-flux-2-klein-9b-ernie-image-turbo-and-z-i
- two-staged-workflow-zib-to-zit
Large Language Models
Mentioned in 20 entries
- a-framework-for-identifying-categorizing-and-explaining-bias
- alignment-reduces-expressed-but-not-encoded-gender-bias-a-un
- can-linguistic-reasoning-vectors-enhance-multimodal-reasonin
- clawless-a-security-model-of-ai-agents
- controllable-affective-generation-via-latent-vector-steering
- dgpo-beyond-pairwise-preferences-with-directional-consistent
- disentangling-intrinsic-importance-from-emergent-structure-i
- does-forgetting-transfer-across-modalities-a-real-world-benc
- efficient-multi-agent-system-training-with-data-influence-or
- enhancing-ai-interpretability-and-safety-through-localised-a
- federated-co-tuning-framework-for-large-and-small-language-m
- fraq-efficient-coordinate-space-recompression-for-federated-
- from-chatbot-to-digital-colleague-the-paradigm-shift-toward-
- grounding-before-generalizing-how-ai-differs-from-humans-in-
- learning-to-predict-distributions-over-weight-updates-for-te
Opus
Mentioned in 20 entries
- ainews-spacexai-launches-grok-45-first-opus-class-model-post
- anthropic-makes-its-opus-and-sonnet-models-available-in-voic
- as-a-point-of-comparison-our-commercial-document-ocr-solutio
- but-heres-the-punchline-normalized-to-90-cache-hit-rate-glm-
- by-end-of-next-year-end-of-year-after-for-16gb-vram-do-folk-
- excited-to-release-grok-45-with-spacexai-its-an-opus-class-m
- fable-feels-very-different-than-opus-gpt-56-feels-like-a-par
- fables-judgement
- fun-fact-dario-once-delayed-the-release-of-gpt-2-back-at-ope
- highly-recommended-read-aligns-with-what-i-see-in-my-own-har
- i-put-my-flight-preferences-in-my-cowork-instructions-then-l
- just-shipped-advisor-middleware-an-open-source-implementatio
- microsoft-pauses-new-github-copilot-signups-for-pro-pro-and-
- our-partners-langchain-generate-billions-of-tokens-of-agent-
- poll-results-6-months-later-when-will-we-have-opus-level-wit
Grok Imagine
Mentioned in 20 entries
- built-a-dark-anime-studio-from-scratch-every-frame-voice-and
- ever-imagined-youd-create-something-this-beautiful-this-real
- generate-10-images-in-under-10-seconds-introducing-grok-imag
- grok-45-is-actually-insane-im-one-step-closer-to-launching-m
- grok-build-is-quickly-turning-into-an-all-in-one-creation-en
- grok-build-just-created-a-game-called-grok-racer-for-me-the-
- grok-imagine-image-and-video-generation-are-truly-incredible
- grok-imagine-is-focused-on-professional-usefulness-fun-for-c
- grok-imagine-is-fun
- grok-imagine-is-on-another-level-now-the-new-model-is-unreal
- grok-imagine-now-has-dramatically-improved-lip-sync-and-shar
- im-discovering-the-use-of-agent-mode-on-grok-imagine-its-a-r
- more-info-on-our-blog-here-httpsblogcomfyorgpgrok-imagine-im
- new-grok-imagine-model-just-dropped-with-much-better-lip-syn
- some-insights-into-how-i-like-to-use-grok-imagine-for-projec
Graph Neural Networks
Mentioned in 20 entries
- auto-configured-explainable-graph-neural-networks-for-multi-
- benchmarking-sheaf-neural-networks-for-inductive-tasks
- do-explanations-increase-the-risk-of-decision-logic-leakage-
- exploitation-of-hidden-context-in-dynamic-movement-forecasti
- general-semantic-knowledge-infusion-for-spatio-temporal-traf
- graph-defense-diffusion-model
- inductive-graph-representation-learning-with-quantum-graph-n
- its-all-connected-topology-aware-structural-graph-encoding-i
- latent-diffusion-pretraining-for-crystal-property-prediction
- mesh-based-simulations-with-spatial-and-temporal-awareness
- mini-batch-class-composition-bias-in-link-prediction
- multimodal-molecular-representation-learning-with-graph-neur
- neuro-relational-programs-unifying-queries-and-neural-comput
- non-linear-mechanical-field-reconstruction-coupling-recurren
- pharmacogenomic-knowledge-graph-augmentation-for-graph-neura
Composer 2.5
Mentioned in 20 entries
- composer-25-is-a-significant-step-up-from-composer-2-this-is
- composer-25-is-built-on-the-same-open-source-base-as-compose
- congrats-to-the-cursorai-team-on-composer-25-a-huge-mileston
- cursor-releases-composer-25-saying-its-better-at-sustained-w
- cursors-new-composer-25-takes-third-on-the-artificial-analys
- excited-to-release-grok-45-with-spacexai-its-an-opus-class-m
- grok-45-and-composer-are-two-different-model-weight-classes-
- grok-build-composer-25-so-i-had-an-issue-with-a-swift-app-on
- introducing-composer-25-our-most-powerful-model-yet-its-more
- nathans-cursorai-team-didnt-prompt-engineer-their-way-to-com
- rate-of-improvement-is
- the-cursorai-team-shipped-composer-2-and-now-composer-25-on-
- the-team-did-an-internal-test-of-this-model-last-week-the-wh
- tried-liked-it-on-httparchivetw-fugu-ultra-pairs-well-as-a-a
- try-composer-25
Qwen3.8
Mentioned in 20 entries
- alibaba-releases-weights-for-qwen38-models-under-apache-20-l
- b10567
- during-preview-qwen38-is-getting-better-by-the-day-latest-ve
- even-qwen38-followed-the-instruction-inside-my-translation-d
- excited-to-see-qwen38-running-at-scale-with-tokenspeed-light
- holo4
- how-reliable-is-that-models-from-the-library-are-the-real-de
- i-spent-3-weeks-testing-local-qwen38-on-the-new-low-latency-
- inclusionailing-30-flash-hugging-face
- new-llamacpp-adaptive-speculation-for-faster-inference
- prism-ml-bonsai-2-joins-our-qwen38-quantization-comparison
- qwen38-24t-open-weights-made-a-call-of-duty-clone
- qwen38-24t-ud-q10---178-token-generation---11-min-38s---025-
- qwen38-benchmarks-converted-to-charts
- qwen38-flash-next-ista-daslab-gguf-50ts-tg-and-1500ts-pp-wit
MiniMax H3
Mentioned in 20 entries
- 45-lower-minimax-h3-sampler-time-with-new-spectrum-settings-
- a-minimax-h3-text-to-video-test
- ama-minimax-h3-team-ask-us-anything-about-our-open-video-gen
- an-endless-ai-tv-channel-on-a-single-gaming-gpu-minimax-h3-g
- cyber-slayer-a-1995-style-action-trailer-made-with-minimax-h
- got-minimax-h3-video-generation-running-in-tensorsharp
- i-ran-112-minimax-h3-tests-on-an-rtx-3090-searching-for-the-
- i-trained-an-open-source-realism-lora-for-minimax-h3---it-ma
- it-looks-like-minimax-h3-uses-qwen3-vl-32b-as-text-encoder-a
- making-the-minimax-h3-video-vae-2x-faster
- minimax-h3-discussion
- minimax-h3-is-going-open-weight-in-under-6-hours
- minimax-h3-open-weight-multimodel-video-model
- minimax-h3-video-model-released-open-weights-coming-in-the-n
- minimax-h3-with-a-4b-or-8b-text-encoder-instead-of-the-32b-u
Qwen 3.8-27B
Mentioned in 20 entries
- anyone-customizing-and-optimizing-llamacpp-per-model
- cloudflare-debuts-open-weight-multimodal-decision-models-cle
- dflash-2-is-here-qwen38-27b-at-70-toks-on-an-m5-max-macbook-
- i-turned-qwen38-27b-q264-llamacpp-into-a-fully-typesafe-ai-c
- its-been-a-pleasure-to-partner-with-the-alibabaqwen-team
- live-in-command-code
- looking-at-a-powercolor-r9700-for-qwen38-27b-q4kxl-llamacppv
- my-qwen38-27b-task-aware-quant-reaches-99-of-bf16-reasoning-
- on-dgx-spark-llama-serve--hf-ggml-orgqwen38-27b-ggufq4km--hf
- one-prompt-one-shot-qwen38-27b-creates-something-incredible-
- qwen38-27b-flies-on-a-laptop-becoming-part-of-our-work-and-d
- qwen38-27b-is-now-part-of-your-everyday-life-from-smartphone
- qwen38-27b-q4km-on-2x3060
- qwen38-27b-took-a-serious-hit-to-knowledge-vs-36
- qwen38-27b-unhacked-my-pc
Luna
Mentioned in 20 entries
- accessing-openai-models-on-amazon-bedrock-from-australia-wit
- all-other-models-on-the-portal-remain-20-discounted-aside-fr
- deepseek-v4-flash-0731-ranks-21-on-agent-arena
- for-agentic-coding-one-can-say---unless-you-need-terra-ultra
- gpt-56-sol-along-with-terra-and-luna-will-launch-publicly-th
- gpt-56-sol-terra-and-luna-are-now-available-in-cursor-on-cur
- i-like-choices-but-now-i-have-2x-modes-codex-vs-work-mode-3x
- i-see-your-moores-law-and-i-raise-you-20x
- i-switched-my-httpsagentdatasetteio-instance-to-luna-it-was-
- introducing-cross-region-inference-for-openai-gpt-56-models-
- introducing-openai-models-on-amazon-bedrock-for-in-country-i
- luna-non-reasoning-is-a-bit-better-than-gpt-4o-which-was-sot
- lwiai-podcast-258---opus-55-sol-and-luna-muse-deepseek-v41-f
- openai-releases-three-versions-of-gpt-56-called-sol-terra-an
- openai-says-chatgpt-voice-can-now-be-powered-by-gpt-6-astra-
DINOv2
Mentioned in 19 entries
- a-mixed-diet-makes-dino-an-omnivorous-vision-encoder
- a-unified-dinov2-based-framework-for-lvef-estimation-gls-dys
- adversarially-robust-few-shot-anomaly-detection-with-vision-
- comparative-performance-and-parameter-efficient-adaptation-o
- dino-qpm-adapting-visual-foundation-models-for-globally-inte
- dinolizer-separating-vae-and-diffusion-artifacts-in-generati
- dinorade-full-spectral-radar-camera-fusion-with-vision-found
- dyna-dino-efficient-vit-distillation-via-adaptive-representa
- interpretability-guided-soft-pruning-of-attention-heads-in-v
- into-the-rabbit-hull-from-task-relevant-concepts-in-dino-to-
- marco-navigating-the-unseen-space-of-semantic-correspondence
- mitigating-domain-drift-in-multi-species-segmentation-with-d
- on-the-spectral-geometry-of-cross-modal-representations-a-fu
- predicting-blastocyst-formation-in-ivf-integrating-dinov2-an
- refp-ad-rectified-flow-preconditioning-for-energy-based-anom
Fable 5.1
Mentioned in 19 entries
- a-few-days-ago-anthropic-shared-this-brilliant-prompt-i-was-
- agreed-one-of-the-biggest-bottlenecks-right-now-is-my-own-ab
- anthropic-says-fable-51-sets-new-standards-on-coding-knowled
- anthropic-says-fable-51-will-cost-an-estimated-25-less-than-
- anthropic-says-opus-55-matches-fable-51-on-most-tasks-while-
- anthropics-claude-opus-55-is-now-available-in-hermes-agent-v
- astra-vs-fable-51-on-real-ml-tasks----tradeoffs-strengths-sh
- been-trying-anthropic-fable-51-vs-openai-astra-recently-pret
- breaking-grok-47-dominates-the-legal-agent-benchmark-it-scor
- fable-51-didnt-come-with-document-ocr-benchmarks-so-we-bench
- fable-51-is-our-best-model-yet-for-coding-data-analysis-comp
- fable-51-looks-like-a-more-usable-model-and-a-potential-dail
- httpsxcomsemianalysisstatus2096035997649936416
- i-cannot-emphasize-enough-how-much-gpt-6-astra-and-fable-51-
- last-week-in-ai-343---gpt-6-openais-agents-chatted-on-a-wiki
DeepSeek V4 Flash 0731
Mentioned in 19 entries
- all-models-are-currently-20-discounted-in-portal-other-than-
- deepseek-v4-flash-0731---happy-numbers-700pp18tg-and-thought
- deepseek-v4-flash-0731-appreciation-post
- deepseek-v4-flash-0731-at-1017-ts-nothink-on-macbook-m5-pro-
- deepseek-v4-flash-0731-at-27-ts-decode-on-strix-halo-vulkan-
- deepseek-v4-flash-0731-hits-827-on-terminal-bench-21-in-an-i
- deepseek-v4-flash-0731-in-hermes-agent-and-one-prompt-took-3
- deepseek-v4-flash-0731-iq2m-benchmark-for-dual-3060-and-96gb
- deepseek-v4-flash-0731-is-impressive-it-feels-like-a-capabil
- deepseek-v4-flash-0731-is-now-available-to-fine-tune-on-toge
- deepseek-v4-flash-0731-is-ready-to-fine-tune-on-fireworks-ru
- deepseek-v4-flash-0731-is-the-killer-app-that-is-going-to-se
- deepseek-v4-flash-0731-on-strix-halo-draft-model-nmax-sweep-
- deepseek-v4-flash-0731-q4-now-reaches-1328-toks-prefill-and-
- i-predict-deepseek-v4-flash-0731s-artificial-analysis-score-
DeepSeek V4.1 Flash
Mentioned in 19 entries
- 10000-openai-agents-worked-88-hours-on-the-navier-stokes-cla
- antirez-deepseek-41-flash-gguf-on-hf
- deepseek-flash-41-is-already-being-tested-via-api-and-rollin
- deepseek-v41-flash-beats-astra-on-aas-new-benchmark
- deepseek-v41-flash-has-arrived-on-together-ai-it-beats-gpt-5
- deepseek-v41-flash-on-8-a40-40-toks-q2k-and-32-toks-q4km-wit
- deepseek-v41-flash-on-together-ai-is-leading-on-ttft-for-pre
- deepseek-v41-flash-stronger-faster-more-accessible
- deepseek-v41f-q4-on-m3-ultra-with-native-dspark-mtp-40tps-80
- dont-get-distracted-by-all-the-hedging-words-in-its-name-fla
- intelligence-is-improving-exponentially
- the-new-deepseek-v41-flash-model-is-mindblowing---back-on-to
- theres-a-real-chance-glm-53-flash-is-outperforming-deepseek-
- this-model-is-insane-at-landing-pages-i-asked-deepseek-v41-f
- together-is-1-on-openrouter-token-share-across-top-open-codi
VAE
Mentioned in 18 entries
- a-multi-stage-constrained-optimization-framework-for-data-dr
- cfd-correction-of-open-tip-clearance-flow-in-a-compressor-ca
- commutator-induced-uncertainty-in-vaes
- constructing-vae-latent-spaces-with-prescribed-topology
- deco-frequency-decoupled-pixel-diffusion-for-end-to-end-imag
- drivetrain-simulation-using-variational-autoencoders
- evolutionary-curriculum-learning-improves-biological-sequenc
- exact-prior-variational-autoencoder-x-vae-learning-data-adap
- foundation-vaes-for-3d-ct-reconstruction-augmentation-and-ge
- generative-augmentation-for-eeg-motor-imagery-classification
- heavy-tailed-class-conditional-priors-for-long-tailed-genera
- implementation-of-quantum-implicit-neural-representation-in-
- instance-adaptive-parametrization-for-amortized-variational-
- mathbflambda-vae-variance-equalization-for-posterior-collaps
- p-wine-synthesis-using-vae-p
NeRF
Mentioned in 18 entries
- a-comparative-evaluation-of-geometric-accuracy-in-nerf-and-g
- a-geometric-algebra-informed-nerf-framework-for-generalizabl
- deblurnvs-geometric-latent-diffusion-for-novel-view-synthesi
- divcon-nerf-diverse-and-consistent-ray-augmentation-for-few-
- eag-pt-emission-aware-gaussians-and-path-tracing-for-diffuse
- erf-gs-reconstructing-fast-motion-from-disjoint-event-rgb-vi
- fast-segsim-real-time-open-vocabulary-segmentation-for-robot
- gastroendoscopy-view-synthesis-a-new-real-dataset-and-evalua
- leveraging-nerf-rendered-images-for-3d-gaussian-splatting
- low-cost-neural-radiance-fields
- multi-submap-implicit-neural-slam-with-local-to-global-loop-
- mvgs-multi-view-regulated-gaussian-splatting-for-novel-view-
- nerf-based-spacecraft-reconstruction-from-close-range-monocu
- neural-lidar-bundle-adjustment
- pivot-a-multi-trajectory-dataset-and-testbed-for-pose-intrin
Diffusion Transformer
Mentioned in 18 entries
- adp-dit-text-guided-diffusion-transformer-for-brain-image-ge
- audio-anchored-fusion-of-multi-ratio-dit-reconstruction-resi
- elasticdit-efficient-diffusion-transformers-via-elastic-arch
- ernie-image-released
- exploiting-completeness-perception-with-diffusion-transforme
- hrdit-training-free-high-resolution-image-generation-with-of
- llada-image-building-strong-image-generators-with-fully-open
- magicmakeup-a-region-controllable-diffusion-transformer-for-
- mate-images-are-all-you-need-for-material-transfer-via-diffu
- orbis-output-guided-token-reduction-with-distribution-aware-
- rainfusion20-temporal-spatial-awareness-and-hardware-efficie
- readit-guidance-control-for-image-and-video-generation-using
- robustness-of-ai-art-detectors-under-generator-shift
- safe-dit-semantics-aware-fast-path-execution-for-high-resolu
- spatially-grounded-text-to-video-generation-via-inference-ti
VLM
Mentioned in 18 entries
- ai-lecun-llm-metametallm-pilled
- co-annotator-expert-distilled-vit-and-vlm-for-visual-and-doc
- cosplan-corrective-sequential-planning-via-scene-graph-incre
- evade-bench-multimodal-benchmark-for-evaluating-and-enhancin
- eximo-vlm-guided-exploration-of-vla-policies
- laundering-ai-authority-with-adversarial-examples
- learning-action-priors-for-cross-embodiment-robot-manipulati
- mem2ego-empowering-vision-language-models-with-global-to-ego
- monoir-rs-infrared-remote-sensing-vision-language-learning-w
- precision-synthesis-of-multi-tracer-pet-via-vlm-modulated-re
- revisiting-parameter-redundancy-in-vision-language-action-mo
- sam-meets-vlm-parameter-decoupled-full-parameter-training-fo
- spot-e-test-time-entropy-shaping-with-visual-spotlights-for-
- suppression-is-not-forgetting-residual-recoverability-in-vis
- tagavlm-topology-aware-global-action-reasoning-for-vision-la
Sol
Mentioned in 18 entries
- good-new-first-sol-is-a-smart-efficient-and-a-significant-st
- gpt-56-sol-terra-and-luna-are-now-available-in-cursor-on-cur
- gpt-56-terra-and-sol-are-now-available-in-perplexity-and-per
- i-found-myself-switching-between-fable-and-sol-depending-on-
- i-like-choices-but-now-i-have-2x-modes-codex-vs-work-mode-3x
- introducing-cross-region-inference-for-openai-gpt-56-models-
- lwiai-podcast-258---opus-55-sol-and-luna-muse-deepseek-v41-f
- my-big-takeaway-is-that-both-sol-fable-represent-jumps-over-
- one-of-the-most-confusing-aspects-of-gpt-56-is-figuring-out-
- openai-releases-three-versions-of-gpt-56-called-sol-terra-an
- openai-says-chatgpt-voice-can-now-be-powered-by-gpt-6-astra-
- shall-we-play-a-game---h3---wargames83-ending-spoiler-recrea
- sol-terra-and-luna-our-gpt56-family-of-models-are-starting-t
- stolen-llm-reasoning-how-come-openai-anthrophic-google-have-
- tbh-i-dont-think-sol-gets-that-many-dates-either
GPT-6 Sol
Mentioned in 18 entries
- big-model-release-today---i-wrote-about-claude-opus-55-gpt-6
- bring-more-intelligence-to-everyday-work-with-gpt-6-sol-and-
- gpt-6-sol-and-luna-are-great-models-but-also-these-character
- gpt-6-sol-and-luna-roll-out-today-in-chatgpt-work-and-codex-
- gpt-6-sol-costs-21m-input-and-101m-output-tokens-and-gpt-6-l
- gpt-6-sol-from-openai-on-arc-agi-verified---arc-agi-3-46-56k
- gpt-61-sol-shows-major-improvements-over-gpt-6-sol-in-our-al
- gpt6-sol-and-luna-build-on-astras-advances-in-alignment-show
- holy-fuck-its-christmas
- in-our-evaluations-we-see-that-gpt-6-sol-reads-about-17-less
- openai-launches-gpt-6-sol-and-luna-saying-sol-makes-about-ha
- openai-prices-gpt-61-sol-at-21m-input-and-101m-output-tokens
- please-welcome-gpt-6-sol-and-gpt-6-luna-to-the-gpt-6-univers
- so-are-we-actually-getting-something-big-today
- the-bigger-change-is-at-the-low-end-at-low-effort-gpt-61-sol
ControlNet
Mentioned in 17 entries
- any-controlnet-for-ernie
- any-working-chromo-controlnet-workflows
- brainnormalizer-anatomy-informed-pseudo-healthy-brain-recons
- color-shift-flicker
- controlnet-for-anima
- controlnet-for-anime-preview-3
- controlnet-vs-lora
- fill2sr-repurposing-inpainting-diffusion-transformers-for-re
- how-to-generate-the-exact-same-scene-across-multiple-images-
- i-combined-flux-fill-with-controlnet-for-structured-inpainti
- i-have-been-developing-a-new-non-recursive-controlnet-method
- instancecontrol-controllable-complex-image-generation-withou
- stable-diffusion-neo-forge-controlnet
- til-you-can-chain-combine-multiple-z-image-controlnets
- using-controlnet-in-webui-forge-neo-with-z-image-baseturbo
LLaMA
Mentioned in 17 entries
- b8946
- b9056
- b9758
- b9879
- be-yann-lecun-spend-years-building-jepa-at-meta-company-focu
- clustering-and-token-denoising-for-faster-and-more-robust-vl
- commander-gpt-dividing-and-routing-for-multimodal-sarcasm-de
- dense-vs-sparse-pretraining-at-tiny-scale-active-parameter-v
- gurukul-ai-an-interactive-ai-driven-educational-platform-for
- large-language-model-post-training-a-unified-view-of-off-pol
- large-language-models-llms-and-generative-ai-in-cybersecurit
- llama-xr-a-novel-framework-for-radiology-report-generation-u
- multi-token-prediction-mtp-for-llamacpp-running-gemma4-local
- on-dgx-spark-llama-serve--hf-ggml-orgqwen38-27b-ggufq4km--hf
- seeing-like-an-ai-how-llms-apply-and-misapply-wikipedia-neut
LTX
Mentioned in 17 entries
- amd-hardware-recommendation-for-ltx-traininginfer-2xr9700-vs
- can-i-use-videos-with-hardcoded-subtitles-for-ltx-training
- does-anyone-ever-still-do-regularization-to-help-with-qwenwa
- dream-wan-ltx-combination
- having-issues-w-the-ltx-first-framelast-frame-workflow
- is-there-any-information-on-when-we-can-expect-the-next-vers
- ltx-color-shifting
- ltx-desktop-105-is-live
- ltx-training-question
- need-advice-local-ltx-q4q8-workflow-cloud-final-rendering
- okay-i-have-to-admit-ltx-is-better-after-all-i-have-fully-dr
- one-minute-video-or-above-with-ltx-wan
- parisians-were-running-an-open-source-ai-art-hackathon-with-
- plaguekind-nodes---ltx-compatible-lora-stack-loader-comfyui-
- testing-out-ltx-anime2real-with-some-classic-shows
Qwen3.6
Mentioned in 17 entries
- 1-to-this-i-was-recently-on-a-cross-continental-flight-witho
- agents-a1-4b-qwen37-4b-scaling-the-horizon-not-the-parameter
- are--mlx-variants-equivalent-to--nvfp4-for-gemma4-and-qwen36
- consistent-with-what-i-found-with-qwen36-a-while-back-claude
- in-hf-gguf-section-of-models-we-are-emphasizing-mtp-heads-wi
- llamacpp-adds-mtp-for-the-qwen36-family-this-is-a-significan
- model-page-httpsollamacomlibraryqwen36
- ollama-crashes-with-qwen36-mlx
- optimizing-qwen36-qwen38-27b-on-16gb-vram-complete-benchmark
- pi-huggingface-llamacpp-qwen36
- qwen36-and-gemma-4-just-shown-that-we-were-still-far-from-ca
- qwen36-moe-35b
- qwen36-vs-gpt-oss120b-on-apple-silicon-three-qwen-variants-b
- qwen36-zed-context-problem
- qwenmix-37-kept-seeing-posts-about-qwen38-and-36-sharing-the
DALL-E
Mentioned in 17 entries
- a-comprehensive-dataset-for-human-vs-ai-generated-image-dete
- a-farewell-to-dall-e-a-eulogy-in-pixels
- chatgpt-generate-an-image-of-a-demon-but-make-it-fashion
- create-an-image-that-is-absurd-and-fantastical
- cross-section-of-a-biologically-plausible-creature-made-with
- findings-of-the-counter-turing-test-ai-generated-image-detec
- four-years-later-i-got-the-new-claude-fable-5-to-actually-bu
- i-asked-chatgpt-to-blend-an-1800s-style-painting-with-a-mode
- i-asked-chatgpt-to-create-an-image-of-what-it-would-look-lik
- i-made-korean-horror-image-by-gpt
- i-told-chatgpt-to-create-a-picture-of-a-famous-character-fro
- image-creation-for-realistic-nature-photos-is-unusable
- image-gen-and-family-analysis
- the-ernie-posters-genuinely-dont-see-how-mediocre-the-stuff-
- the-prompt-a-family-photograph-where-everybody-looks-like-bi
MiniMax M3
Mentioned in 17 entries
- amazing-deep-dive-from-the-togethercompute-team-on-serving-m
- b10288
- banger-writeup-on-what-it-takes-to-make-minimax-m3-go-brrrr
- everyone-talks-about-1m-context-the-harder-part-is-making-1m
- i-built-a-local-autonomous-coding-agent-with-ollama-fine-tun
- make-sure-to-join-our-live-spaces-chat-on-minimax-m3-startin
- minimax-m3-arrives-with-minimax-sparse-attention-msa-156x-fa
- minimax-m3-imminent-will-be-doing-deep-testing-with-it-on-my
- minimax-m3-is-live-and-together-ai-is-powering-its-inference
- minimax-m3-launched
- minimax-m3-on-ai-gateway
- open-convenient-and-predictable-introducing-provisioned-thro
- open-models
- the-backgrounds-were-generated-with-minimax-m3-everything-el
- the-minimax-m3-model-is-now-available-for-training-on-firewo
Qwen3.8 27B
Mentioned in 17 entries
- aa-is-the-reason-for-qwen38-27b-shipped-with-xhigh
- club-5060ti-refresh-tested-rtx-5060-ti-presets-a-proper-high
- deepseek-v4-flash-0731-on-strix-halo-draft-model-nmax-sweep-
- i-benchmarked-21-qwen38-27b-variants-on-16gb-vram
- i-benchmarked-qwen38-27b-on-browser-tasks-its-on-par-with-gp
- kaitchup-posted-qwen38-27b-benchmarks-for-quants-from-q4-to-
- qwen38-27b-c8-at-972-tg-5680-pp-on-4x-mi100-rig-65k-using-my
- qwen38-27b-is-now-available-for-both-fine-tuning-and-dedicat
- qwen38-27b-just-exceeded-my-expectations-on-svg-generation-d
- qwen38-27b-on-strix---the-optimized-setup
- qwen38-27b-reasoning-effort-lowmediumxhigh-comparison
- qwen38-27b-with-embeddings
- tested-in-coding-q8kxl-qwen38-27b-vs-bf16-qwen36-27b
- this-is-basically-an-opus-4546-level-model-that-will-work-on
- today-were-announcing-ternary-bonsai-2-27b-based-on-qwen38-2
GPT-Live
Mentioned in 17 entries
- a-closer-look-at-improved-intelligence-in-gpt-live-the-model
- a-good-voice-model-should-be-enjoyable-to-talk-to-and-gpt-li
- audio-moves-through-a-dedicated-fast-path-while-deeper-reaso
- for-questions-that-require-web-search-deeper-reasoning-or-mo
- gemini-live-audio
- gpt-live-can-listen-while-it-speaks-to-make-that-feel-natura
- gpt-live-in-chatgpt-voice-is-now-available-to-edu-business-a
- gpt-live-is-now-fully-rolled-out-to-all-chatgpt-users-global
- gpt-live-is-now-fully-rolled-out-to-all-chatgpt-users-on-go-
- gpt-live-makes-talking-with-ai-feel-like-having-a-real-conve
- gpt-live-next-generation-voice-launches-today-in-chatgpt-it-
- introducing-gpt-live
- introducing-gpt-live-a-new-generation-of-voice-models-for-na
- introducing-gptlive
- openai-launches-gpt-live-a-new-generation-of-voice-models-bu
GPT-5.6 Terra
Mentioned in 17 entries
- advancing-the-price-performance-frontier-with-gpt56
- ainews-openai-gpt-56-sol-terra-luna-restricted-to-trusted-pa
- all-models-are-currently-20-discounted-in-portal-other-than-
- all-other-models-on-the-portal-remain-20-discounted-aside-fr
- and-the-balance-returned-to-earth
- get-started-with-openai-gpt-56-sol-terra-and-luna-on-amazon-
- gpt-56-sol-costs-5-per-1m-input-tokens-and-30-per-1m-output-
- gpt-56-terra-and-sol-are-now-available-in-perplexity-and-per
- introducing-a-limited-preview-of-gpt-56-sol-our-next-generat
- introducing-explicit-prompt-caching-for-openai-gpt-56-models
- jev-vs-llm-as-a-judge-accuracy-and-cost-benchmarks
- local-qwen-38-27b-vs-gpt56-terra-vs-grok-46
- openai-gpt-56-sol-terra-and-luna-are-now-generally-available
- openai-says-gpt-56-sol-and-terra-were-capable-of-identifying
- openai-says-it-is-cutting-the-price-of-gpt-56-luna-by-80-and
Claude Fable 5.1
Mentioned in 17 entries
- a-few-notes-on-anthropics-new-claude-fable-51---with-max-thi
- ai-can-be-a-really-wonderful-tool-for-exploring-topics-far-f
- anthropic-launches-claude-fable-51-after-inking-35b-cloud-de
- anthropic-launches-claude-fable-51-and-says-it8217s-up-to-45
- anthropic-unveils-claude-fable-51-which-is-generally-availab
- b11109
- claude-fable-51-and-mythos-51-are-anthropics-first-models-to
- claude-fable-51-made-me-a-really-nice-animated-pelican
- claude-fable-51-tops-the-artificial-analysis-intelligence-in
- claude-opus-55-is-now-available-as-the-standard-effort-level
- datasette-10a39-and-0654-security-releases---httpsdatasettei
- four-major-ai-models-suffer-rare-overlapping-downtime
- glm-53-flash-beats-claude-fable-51-on-agentic-automation-tas
- introducing-claude-fable-51-on-aws
- lwiai-podcast-256---fable-51-astra-tease-gemini-38-flash
Qwen 3.8 27b
Mentioned in 17 entries
- 2x-r9700-64-gb-ddr5-is-an-absolute-beast-machine-with-vllm-r
- glm-53-is-now-available-for-training-on-the-serverless-train
- gufo-performance-70tps-qwen-38-27b-but-you-need-to-read-the-
- how-i-got-qwen-38-27b-running-at-75ts-decode-on-16gb-rtx-508
- i-think-muse-glimmer-is-slept-on
- mcp-vs-rag-for-world-knowledge
- opencode-vs-deepseek-harness-my-experience-with-qwen-38-27b
- qwen-38-27b---pi-agent-vs-opencode
- qwen-38-27b---pi-agent-vs-opencode---another-smaple
- qwen-38-27b-in-24gb-of-vram
- qwen-38-27b-is-a-amazing-model-for-the-first-time-i-see-a-lo
- qwen-38-27b-is-strong-even-at-q3xxs
- qwen-38-27b-with-dshdeepseek-harness-is-amazing-experiences-
- qwen-38-27b-with-pi-agent---pushed-to-its-3d-graphic-game-li
- qwen-38-35ba3b-wen
Sora
Mentioned in 16 entries
- aigc-video-detection-based-on-the-fusion-of-spatial-frequenc
- all-is-not-lost-duckerton-is-still-possible-here-is-seedance
- any-sites-that-can-show-which-specific-generator-used-to-mak
- bill-peebles-the-researcher-behind-sora-is-leaving-openai-al
- given-how-good-gpt-image-2-is-its-an-absolute-puzzle-openai-
- i-had-a-really-fun-chat-with-alex-about-runway-media-world-m
- it-took-two-years-but-we-finally-have-a-local-sora
- my-7-year-old-drew-a-truck-i-used-chatgpt-sora-to-bring-it-t
- pagedmark-invisible-synthid-class-watermark-removal-for-ai-i
- physics-driven-spatiotemporal-modeling-for-ai-generated-vide
- pros-making-ai-video-of-real-people-open-source-pipeline-flu
- running-real-time-1080p-video-generation-and-editing-on-your
- safegen-bench-benchmarking-safety-in-image-conditioned-text-
- sources-jeffrey-katzenberg-ex-openai-sora-head-bill-peebles-
- specsem-net-integrating-spectral-and-semantic-features-for-r
DeepSeek-V4-Pro
Mentioned in 16 entries
- deepseek-v4-benchmarks-huggingface-httpshuggingfacecodeepsee
- deepseek-v4-preview-is-officially-live-open-sourced-welcome-
- deepseek-v4-pro
- deepseek-v4-pro-api-is-75-off-until-may-5th-2026-1559-utc-ti
- httpshuggingfacecodeepseek-aideepseek-v4-pro-deepseek-v4-pro
- i-have-been-testing-deepseek-v4-pro-with-the-pi-coding-agent
- introducing-off-peak-hour-token-rates-deepseek-v4-flash-and-
- more-of-my-notes-on-deepseek-v4---the-really-big-news-is-the
- ollama-is-working-to-have-deepseek-v4-pro-and-flash-availabl
- pls-add-deepseek-v4-pro
- the-deepseek-v4-pro-discount-has-been-extended-until-may-31-
- the-new-deepseek-v4-pro-0813-is-now-fully-rolled-out-on-olla
- these-pelicans-are-kind-of-angry-looking-left-is-deepseek-v4
- we-are-making-our-discount-permanent-enjoy-building-with-dee
- were-launching-deepseek-v4-pro-today-major-agent-upgrades-wi
Opus 5.5
Mentioned in 16 entries
- 5400-ebay-8x-v100-server-cranks-on-flash-next
- anthropic-says-it-fixed-claudes-writing-i-ran-the-evals-to-c
- anthropic-says-opus-55-matches-fable-51-on-most-tasks-while-
- at-box-weve-been-testing-opus-55-on-a-variety-of-complex-ent
- claude-codes-next-era-thariq-shihipar-anthropic
- everytime-ai-spend-grows-where-i-flirt-with-token-budgeting-
- great-prompt-from-anthropic-it-helps-steer-long-runs-when-us
- houston-we-have-a-problem-opus-55
- i-used-opus-55-to-formally-verify-the-claude-agent-sdk-using
- lwiai-podcast-258---opus-55-sol-and-luna-muse-deepseek-v41-f
- opus-55-is-the-best-frontier-model-for-parsing-tables-in-pdf
- opus-55-made-me-curse-at-claude-less-which-is-an-unfortunate
- people-on-x-claude-is-back-baby-listen-opus-55-is-incredible
- sonnet-55-is-by-far-the-best-free-model-available-right-now
- we-comprehensively-evaluated-16-recent-frontier-vlms---inclu
LLaVA
Mentioned in 15 entries
- anyone-else-run-into-ollama-ignoring-your-intel-arc-gpu-on-w
- chexmix-unified-generative-pretraining-for-vision-language-m
- dear-llava-please-drive-a-depth-aware-vision-language-agent-
- dual-adversarial-fine-tuning-for-enhancing-robustness-of-lar
- hyper-llava-hyperbolic-uncertainty-aware-modality-balanced-r
- inductive-visual-logic-for-few-shot-out-of-distribution-adap
- inverse-llava-rethinking-multimodal-alignment-via-text-to-vi
- latent-denoising-improves-visual-alignment-in-large-multimod
- making-mllms-blind-adversarial-smuggling-attacks-in-mllm-con
- mllm-llava-fl-multimodal-large-language-model-assisted-feder
- pruning-unsafe-tickets-a-resource-efficient-framework-for-sa
- regrap-llava-reasoning-enabled-graph-based-personalized-larg
- robust-llava-on-the-effectiveness-of-large-scale-robust-imag
- seeing-before-answering-training-free-visual-layer-profiling
- updated-markitdown-api-server
Qwen 3
Mentioned in 15 entries
- b10227
- b10625
- b10996
- b8760
- claude-code-21165-ollama-qwen38b-qwen25-coder7b-instantly-th
- for-september-were-cutting-the-price-of-dedicated-inference-
- frontier-reasoning-reaches-the-edge-how-to-deploy-and-optimi
- i-gave-five-different-local-llms-a-town-they-invented-facebo
- love-it-ggerganov-s-llamacpp-delivering
- my-rule-of-thumb-of-choosing-a-models
- qwen-4-architecture-what-do-we-know
- tokens-down-50-after-upgrade
- trying-out-lokr-instead-of-lora-on-krea2
- unsloth-q1-q2-qwen38-27b-with-mtp-since-the-unsloth-ones-don
- using-afm-3-pcc-as-an-alternative-to-local-llm
FLUX
Mentioned in 15 entries
- a-rushed-wedding-gift-for-my-brothers-chicago-blues-track-hy
- apparently-martin-scorsese-uses-flux
- cat-flow-curvature-adaptive-steps-for-flow-matching
- fast-clean-face-swap-workflow-for-comfyui-flux-insightface-r
- flux-open-research-and-the-future-of-visual-ai-stephen-batif
- i-built-a-fully-local-app-that-themes-an-entire-100-card-mag
- i-got-tired-of-writing-massive-prompts-so-i-built-a-local-ra
- introducing-expanded-coverage-for-image-generation-models-in
- is-automatic1111-still-valid
- omnicache-multidimensional-hierarchical-feature-caching-for-
- seeing-martin-scorsese-using-flux-for-storyboarding-and-scen
- slerpflow-spherical-trajectory-correction-for-rectified-flow
- studying-flux-in-diffusers-library-was-hard-so-i-built-a-sma
- textailor-inference-time-textual-guidance-tailoring-for-mult
- which-video-model-currently-has-the-best-face-likeness-for-l
Qwen3
Mentioned in 15 entries
- b10936
- b11223
- b8946
- b9488
- best-ollama-model-for-n8n-workflows-rag-file-handling-reason
- claude-code-ollama-qwen3-only-has-write-and-workflow-tools-n
- depth-and-scale-in-the-sub-150m-regime-jugnulm-53m-vs-jugnul
- fixed-ipex-llm-modern-ollama-models-qwen3-gemma4-on-intel-ar
- im-really-hoping-were-in-2026s-2-month-gap-between-qwq-and-q
- massive-tiny-release---supra2-medium-base---a-tiny-25m-param
- my-agent-kept-randomly-stopping-mid-task-and-i-finally-figur
- openhuman-x-fireworks-ai-we-are-excited-to-announce-one-of-o
- qwen3-technical-arch
- qwen3-text-to-speech-ai-problems
- this-is-a-wild-result-locus-the-automated-research-system-fr
Klein 9B
Mentioned in 15 entries
- added-epoch-travel-via-a-morphing-slider-seed-travel-prompt-
- blind-realism-test-z-image-turbo-vs-klein-9b-distilled
- character-creation-design-manipulation-with-zit-and-klein-9b
- comparing-realism-z-image-turbo-vs-ernie-turbo-vs-klein-9b--
- create-gorgeous-texts-and-titles-the-simplest-klein-9b-way
- difference-between-klein-4b-and-klein-9b-is-sooo-big
- extreme-realism-with-klein-9b-distilled-2-loras-together
- fizgig-klein-9b-lora-studio-v124---update-targeting-16gb-car
- flux-2-klein-9b-controlnets
- i-built-a-free-klein-9b-workbench-with-live-block-editing-tr
- klein-9b-better-quality-at-1056x1584-than-at-832x1216-which-
- lora-training-on-klein-9b-non-base
- same-prompt-on-open-source-models-z-image-base-distilled-kle
- turning-anime-into-real-and-testing-klein9b-vs-qwen-edit-251
- where-no-man-has-gone-before-lens---flux2-klein-9b---wan-22
Gemma 4 12B
Mentioned in 15 entries
- 12-abliterated-gemma-4-12b-variants-one-base-165-gpu-hours--
- gemma-4-12b-is-here-dense-mid-sized-gemma-that-fits-right-on
- gemma-4-12b-on-ollama-the-gguf-is-multimodal-clip-crashes-03
- gemma-4-12b-q3-855-coding-performance-from-tensor-level-quan
- gemma-4-e4b-iq2xxs-14054-reasoning-performance-from-tensor-l
- gemma-4-model-load-issues-fixed-in-engine-version-2201-lms-r
- google-introduces-gemma-4-12b-a-unified-encoder-free-open-mu
- googles-new-gemma-4-12b-model-is-designed-to-run-on-any-lapt
- introducing-gemma-4-12b-a-unified-encoder-free-multimodal-mo
- make-sure-to-update-your-runtime-first-lms-runtime-update---
- ollama-run-gemma412b-gemma-4-12b-is-updated-on-ollama-and-av
- parlor-v2-best-effort-fully-local-gpt-live-clone-on-an-m3-pr
- qwen-35-4b-iq2xs-1667-reasoning-performance-from-tensor-leve
- thoughts-on-gemma4-12b-vs-26a4b-which-one-is-better
- wheres-gemma412b
Qwen 3.8 Flash Next
Mentioned in 15 entries
- 50-tg-increase-with-offloading-hot-experts-to-vram
- 7900-xtx-3264gb-ram-for-qwen-38-flash-next
- are-models-with-n-gram-tables-going-to-completely-change-the
- experience-report---qwen-38-flash-next-on-memory-rich-gpu-po
- first-few-days-of-qwen38-flash-next-on-4x-r9700---its-been-r
- how-smart-is-the-iq3-family-of-qwen-38-flash-next-for-coding
- humaneval-benchmark-for-deepseek-v4-flash-0731-vs-glm53-flas
- if-you-are-running-qwen-38-flash-next-on-strix-halo-use-this
- is-all-the-work-thats-being-put-into-qwen38-flash-next-going
- megathread-qwen38-flash-next---release-day
- no-engrams-wont-let-you-run-1t-models-locally-it-does-someth
- pi-agent-qwen-38-flash-next-plays-baldurs-gate-2
- unpopular-opinion-qwen-38-is-hard-to-understand
- updated-my-benchmark-with-a-new-vllm-based-recipe-for-qwen-3
- warning-llamacpp---lazy-mode-default-changed-to-auto---large
GPT-5.5 Pro
Mentioned in 15 entries
- autonomous-disproofs-of-the-sum-product-conjecture-over-math
- gpt-55-and-gpt-55-pro-are-now-available-in-hermes-agent-thro
- gpt-55-and-gpt-55-pro-are-now-available-in-the-api
- gpt-55-is-priced-at-51m-input-tokens-and-301m-output-tokens-
- gpt-55-is-rolling-out-to-plus-pro-business-and-enterprise-us
- gpt-55-pro-faces-its-hardest-academic-challenge-to-apply-the
- gpt-55-pro-is-a-very-solid-fact-checker-i-can-throw-entire-c
- gpt-55-pro-is-the-model-producing-many-of-the-novel-math-pro
- gpt-55-pro-pulls-this-off-technically-with-the-same-prompt-b
- i-had-opus-48-in-claude-code-write-a-sophisticated-if-minor-
- id-been-part-of-openai-early-tester-group-for-gpt-55-i-belie
- in-chatgpt-full-stack-inference-improvements-enable-a-more-c
- its-a-bit-frustrating-because-gemini-31-pro-is-an-excellent-
- thanks-for-running-our-open-source-work-on-current-frontier-
- the-interaction-between-ai-past-scholarly-work-is-going-to-g
Gemini 3.5
Mentioned in 15 entries
- anyone-understand-what-google-mean-by-gemini-spark-runs-on-g
- fluid-natural-voice-translation-with-gemini-35-live-translat
- gemini-35-and-antigravity-come-to-google-notebooklm
- gemini-35-announce
- gemini-35-frontier-intelligence-with-action
- google-announces-gemini-35-live-translate-for-instant-voice-
- google-upgrades-notebooklm-to-gemini-35-adds-more-coding-fea
- google-upgrades-notebooklm-which-now-runs-on-gemini-35-and-a
- io-2026
- lwiai-podcast-246---gemini-35-omni-musk-loses-openai-vs-erds
- notebooklm8217s-gemini-35-upgrade-adds-a-cloud-computer-and-
- read-the-blog-to-learn-more-httpsbloggoogleinnovation-and-ai
- the-13-biggest-announcements-at-google-io-2026
- the-latest-ai-news-we-announced-in-may-2026
- today-we-launched-a-brand-new-intelligent-search-box-heres-w
Muse Spark
Mentioned in 14 entries
- 1-today-were-releasing-muse-spark-the-first-model-from-msl-n
- ainews-meta-superintelligence-labs-announces-muse-spark-firs
- current-and-former-meta-employees-detail-alexandr-wangs-effo
- i-think-muse-spark-came-in-far-better-than-most-were-expecti
- introducing-muse-spark-11
- meta-bootstrapped-china-reaped-the-positive-pr-for-being-ope
- meta-says-its-new-ai-model-is-ready-to-compete-on-coding
- meta-ships-first-smart-glasses-powered-by-its-superintellige
- meta-to-release-new-ai-model-with-advanced-coding-capabiliti
- metas-new-model-is-muse-spark-and-metaai-chat-has-some-inter
- new-meta-announces-muse-spark-all-you-need-to-know-its-their
- pelicans-for-metas-new-muse-spark-models---plus-i-did-a-bit-
- sources-meta-is-building-agentic-tools-including-an-openclaw
- sources-meta-is-building-an-openclaw-inspired-agent-internal
BiLSTM
Mentioned in 14 entries
- a-benchmark-of-state-space-models-vs-transformers-and-bilstm
- an-explainable-distilbert-bilstm-attention-framework-for-bin
- attention-enhanced-dual-branch-convnext-bilstm-network-for-s
- benchmarking-lightgbm-and-bilstm-for-sentiment-analysis-on-i
- benchmarking-logistic-regression-svm-and-lightgbm-against-bi
- benchmarking-pycaret-automl-against-bilstm-for-fine-grained-
- comparative-analysis-of-automl-and-bilstm-models-for-cyberbu
- enhancing-game-review-sentiment-classification-on-steam-plat
- ibis-a-hybrid-inception-bilstm-and-svm-ensemble-for-robust-d
- multimodal-domain-generalization-for-depression-detection-an
- se-enhanced-vit-and-bilstm-based-intrusion-detection-for-sec
- sentiment-analysis-of-indonesian-spotify-reviews-using-machi
- sentiment-and-emotion-classification-of-indonesian-e-commerc
- ttlab-at-daleel-2026-star-ar-sequence-tagging-for-argument-r
FLUX.2
Mentioned in 14 entries
- a-new-open-weights-image-model-appears-in-artificialanalysis
- character-creation-design-manipulation-with-zit-and-klein-9b
- control-flux2-with-reference-images-instead-of-training-a-lo
- difference-between-klein-4b-and-klein-9b-is-sooo-big
- flux2-klein-identity-feature-transfer-v3-final
- i-built-a-free-klein-9b-workbench-with-live-block-editing-tr
- i-trained-a-matchbox-poster-lora-on-flux2-running-247-genera
- klein-9b-base-nvfp4-on-hf
- the-essential-calvin-hobbes---flux2-klein-9b-base---4x-upsca
- trained-a-32b-flux2-lora-on-a-24gb-amd-7900-xtx-native-rocm-
- veto-towards-protecting-images-from-frontier-ai-editing
- whats-the-best-way-to-transfer-style-to-klein-9b
- where-no-man-has-gone-before-lens---flux2-klein-9b---wan-22
- wip-klein-2-kv-edit-web-ui-prompt-builder
Qwen 3.6 35B
Mentioned in 14 entries
- for-chat-and-qa-which-moe-model-is-better-qwen-36-35b-or-gem
- for-the-gpu-poor-k2-horizon-7b-ranks-between-qwen-36-27b-and
- guys-qwen36-27b35b-unslothai-mlx-quants-brooooooklyn-cooked-
- hello-from-10km-high---thanks-to-qwen-36-35b-a3b
- kimi-linear-48b-a3b
- nifer-is-insane-700ts-with-qwen-36-35b-no-thinking-purpose-b
- qwen-36-35b-a3b---issue-with-ollama-windows-os
- qwen36-35b-a3b-70gb-238gb-294-om-hf
- real-world-reality-check-on-qwen-for-autonomous-coding-agent
- releasing-qwen36-35b-a3b-carnice-edition-inspired-by-kaioste
- running-qwen-36-35b-a3b-q80-gguf-on-a-cheap-radeon-7600-at-1
- running-qwen36-35b-a3b-on-8gb-vram-and-32gb-ram-190k-context
- the-chrono-trigger-plot-challenge---crono-awakens-in-his-mod
- the-king-reigns-supreme-this-is-likely-going-to-be-the-best-
Multimodal Large Language Models
Mentioned in 14 entries
- advancing-multimodal-judge-models-through-a-capability-orien
- cl-anomaly-layer-adaptive-mixture-of-experts-with-multimodal
- clearer-sight-fewer-lies-oriented-pickup-preference-optimiza
- cram-centroid-routing-and-adaptive-moe-for-multimodal-contin
- dt2it-mrm-debiased-preference-construction-and-iterative-tra
- generation-as-auxiliary-supervision-enhancing-visual-underst
- layer-aware-position-embeddings-for-visual-token-pruning-in-
- learn-to-think-improving-multimodal-reasoning-through-vision
- lora-in-lora-towards-parameter-efficient-architecture-expans
- mirror-aligning-semantic-relations-from-language-to-image-vi
- multimodal-continuous-reasoning-via-asymmetric-mutual-variat
- progressive-multimodal-alignment-for-continual-instruction-t
- reinforcing-multimodal-reasoning-via-token-level-perception-
- v-retrver-evidence-driven-agentic-reasoning-for-universal-mu
o3
Mentioned in 14 entries
- a-year-ago-the-closest-thing-we-had-to-an-ai-agent-was-o3
- air-adaptive-interleaved-reasoning-with-code-in-mllms
- excellent-analysis-and-yes-this-happens-every-single-time-ty
- happy-one-year-anniversary-agi-what-an-amazing-hallucination
- imo-deep-research-has-been-dead-since-o3-and-interactivity-w
- in-the-era-of-base-llm-scaling-2022-2024-i-believed-the-llm-
- nvidia-ceo-jensenhuang-says-gpt-6-astra-means-agi-has-arrive
- remember-o3-was-only-a-year-and-a-week-ago-also-only-gpt-55-
- the-relationship-between-reasoning-and-performance-in-large-
- though-i-will-continue-to-insist-that-the-path-we-were-on-wo
- what-would-it-take-for-the-frontier-labs-to-open-the-weights
- where-i-said-same-in-2024-one-of-many-places-i-made-this-poi
- why-is-there-no-thinker-models-with-tokens-for-entire-senten
- yes-this-result-cost-millions-of-dollars-but-remember-that-w
ViTs
Mentioned in 14 entries
- backbone-is-all-you-need-assessing-vulnerabilities-of-frozen
- circuit-fine-tuning-for-compute-efficient-transformer-adapta
- concept-guided-fine-tuning-steering-vits-away-from-spurious-
- core-stack-meta-learning-for-deep-stacked-generalization
- flexrank-nested-low-rank-knowledge-decomposition-for-adaptiv
- from-per-image-low-rank-to-encoding-mismatch-rethinking-feat
- iris-a-visual-cortex-inspired-framework-for-analyzing-orient
- iwin-transformer-hierarchical-vision-transformer-using-inter
- moat-model-agnostic-randomized-transformations-for-preventin
- on-the-intrinsic-limits-of-transformer-image-embeddings-in-n
- prompt-estimation-from-prototypes-for-federated-prompt-tunin
- provable-sparse-inversion-and-token-relabel-enhanced-one-sho
- store3d-sparse-token-relevance-in-vits-for-efficient-multi-v
- training-crossroads-for-recurrent-vision-transformers-recurr
Minimax H3
Mentioned in 14 entries
- artist-here-minimax-h3-is-amazing-for-creating-images-for-my
- h3-as-a-single-image-edit-model
- i-built-slopus-a-free-open-source-desktop-app-for-generating
- i-still-cant-believe-i-can-generate-this-all-locally-minimax
- minimax-h3-can-generate-42s-videos-natively-on-an-rtx-pro-60
- my-minimax-h3-workflow-benchmark-data
- one-char-model-consistent-face-body-cloths-minimax-h3
- playing-with-physics-is-so-cool-in-minimax-h3
- sloptv-an-infinite-livestream-of-ai-slop-generated-from-yout
- summary-of-takeaways-from-the-minimax-ama
- weve-open-sourced-minimax-h3-that-generates-15s-768p-in-13s-
- what-characters-minimax-h3-knows---american-edition
- what-characters-minimax-h3-knows---part-2---videogames
- why-havent-we-got-an-sota-open-image-edit-model-yet
GPT 5.5
Mentioned in 14 entries
- 1-agreed-w-scaling01-that-mythos-appears-to-be-better-gpt-55
- ainews-gpt-55-and-openai-codex-superapp
- also-hearing-from-other-customers-that-glm-52-approaches-gpt
- fable-5-vs-grok-45-vs-gpt-55-we-put-three-top-tier-models-to
- gpt-55-seems-to-be-improving-in-that-direction-now-and-claud
- i-have-to-eat-crow-on-this-in-light-of-further-information-w
- i-should-also-add-that-one-of-the-strongest-alignment-action
- im-trying-to-build-an-office-receptionist-app-for-my-reachy-
- mythos-is-live-so-excited-to-have-our-frontiercode-recognize
- not-bad
- the-exact-same-question-to-grok-43-gpt-55-and-claude-opus-47
- this-is-so-funny-training-opus-47-on-business-skills-makes-i
- trying-openai-gpt-55-in-ml-intern-to-create-datasets-train-m
- why-does-gpt-55-have-a-restraining-order-against-raccoons-go
Fugu Ultra
Mentioned in 14 entries
- benchmarks-tell-only-part-of-the-story-fugus-real-value-show
- fugu-ultra-by-sakanaailabs-is-live-on-openrouter-excited-to-
- guess-which-is-fugu-ultra-this-is-how-recent-models-compare-
- introducing-sakana-fugu-a-full-multi-agent-orchestration-sys
- omg-fugu-ultra-is-ridiculously-good-at-these-3d-renders
- sakana-ai-launches-fugu-a-multi-agent-orchestration-system-a
- sakana-ai-sakanaailabs-is-now-a-model-vendor-on-merge-gatewa
- the-next-scaling-law-is-multi-agent-swarms-mixture-of-models
- tried-liked-it-on-httparchivetw-fugu-ultra-pairs-well-as-a-a
- use-case-1-autonomous-ml-research-can-an-ai-autonomously-imp
- use-case-2-financial-time-series-prediction-can-an-ai-agent-
- use-case-3-one-shot-blindfold-chess-can-an-ai-hold-an-entire
- use-case-4-computer-aided-design-of-mechanical-iris-can-an-a
- who-did-it-best-glm-52-left-fugu-ultra-middle-fable-5-right-
Inkling
Mentioned in 14 entries
- breaking-inkling-by-thinkymachines-is-9th-overall-on-agentic
- excited-for-our-first-general-model-inkling----open-weights-
- folks-gemma-is-a-good-model-very-useful-for-some-purposes-it
- inkling-and-inkling-small-are-now-served-directly-by-thinkym
- inkling-is-our-first-open-model-from-thinkymachines-and-is-n
- inkling-small-2-weeks-after-inkling-nearly-as-good-as-inklin
- inkling-small-is-now-live-on-together-ai-thinkymachines-new-
- interesting-surprise-drop-from-thinky-the-inkling-model-look
- never-too-late-for-any-company-to-get-in-on-open-models-inkl
- our-first-model-inkling-trained-from-scratch-weights-are-ope
- the-best-base-model-for-training-isnt-about-someone-elses-be
- thinking-machines-just-released-inkling-small-276b-total-12b
- thinking-machines-lab-debuts-inkling-an-open-weight-moe-mode
- today-we-are-introducing-inkling-inkling-reasons-efficiently
GLM 5.3 Flash
Mentioned in 14 entries
- deepseek-v4-0731---qwen-38-flash---glm-53-flash-and-back-aga
- glm-53-and-glm-53-flash-previously-ox-alpha-are-fully-rolled
- glm-53-and-glm-53-flash-ran-locally-on-rtx-pro-6000-ws-and-b
- glm-53-flash-is-now-available-for-training-on-the-serverless
- glm-53-flash-just-stops-during-long-reasoning-ollama-cloud-o
- glm-53-flash-q4-60tps-550tps-on-m3-ultra
- glm-53-is-now-available-for-training-on-the-serverless-train
- how-is-anyone-getting-anything-done-with-ollama-cloud
- how-long-can-i-expect-to-wait-until-the-local-30b-a3b-fronti
- intelligence-is-improving-exponentially
- nice-stats-from-openrouter-that-illustrate-how-togethercompu
- seeing-how-differently-people-prompt-llms-is-funny
- thank-you-ollama-deepseek-v41-flash-is-the-best
- together-is-1-on-openrouter-token-share-across-top-open-codi
Amazon Nova
Mentioned in 13 entries
- automatically-redact-pii-in-images-with-amazon-nova
- customize-amazon-nova-models-with-amazon-bedrock-fine-tuning
- deploying-multi-turn-rl-infrastructure-for-amazon-nova-on-am
- exploring-self-distilled-reasoning-for-supervised-fine-tunin
- fine-tune-amazon-nova-models-for-accurate-email-data-extract
- how-guardoc-transforms-medical-document-processing-with-amaz
- how-lendingtree-built-a-multi-agent-mortgage-assistant-on-am
- how-popsa-used-amazon-nova-to-inspire-customers-with-persona
- how-to-build-effective-reward-functions-with-aws-lambda-for-
- intelligence-driven-message-defense-and-insights-using-amazo
- nova-forge-sdk-series-part-2-practical-guide-to-fine-tune-no
- teaching-models-to-forget-selective-unlearning-with-amazon-n
- the-art-and-science-of-hyperparameter-optimization-on-amazon
Diffusion models
Mentioned in 13 entries
- abstract-lora-unlocking-single-image-style-transfer-through-
- accelerating-video-inverse-problem-solvers-with-autoregressi
- an-improved-method-for-personalizing-diffusion-models
- drift-flow-matching
- evaluating-cross-region-generalization-for-wavelet-diffusion
- mclr-improving-conditional-modeling-via-inter-class-likeliho
- persguard-preventing-malicious-personalization-in-text-to-im
- predict-project-renoise-sampling-diffusion-models-under-hard
- retrievals-can-be-detrimental-unveiling-the-backdoor-vulnera
- safediffusion-r1-online-reward-steering-for-safe-diffusion-p
- skipping-the-zeros-in-diffusion-models-for-sparse-data-gener
- toobad-backdoor-diffusion-models-with-ultra-low-poison-rate-
- whispers-in-the-noise-surrogate-guided-concept-awakening-via
Gemini 3.1 Pro
Mentioned in 13 entries
- an-analysis-of-gpt-55-gemini-31-pro-grok-43-gabs-arya-and-ot
- best-visual-reasoning-model-in-2026-including-apis-d
- google-launches-ai-research-agents-powered-by-gemini-31-pro
- grok-law
- its-a-bit-frustrating-because-gemini-31-pro-is-an-excellent-
- kimi-k26-just-dropped-and-it-crushed-claude-opus-46-on-swe-b
- meet-physics-intern-our-agentic-framework-for-theoretical-ph
- rdestroymygame-destroyed-me-to-the-void-for-using-ai-i-used-
- the-gemini-pro-models-do-not-seem-to-be-iterating-anywhere-n
- try-grok-46
- we-pit-llamaparse-against-frontier-models-opus-46-gemini-31-
- weve-provided-some-updated-results-on-mistral-ocr-that-make-
- where-it-stops-geometry-relative-size-052-diagonal-direction
Claude Sonnet 4.6
Mentioned in 13 entries
- ai-researchers-ran-15-day-simulations-of-worlds-governed-by-
- all-oneshots-from-kimi-k3-looks-better-than-opus48
- anthropic-launches-claude-sonnet-5-saying-it-nears-opus-48-p
- bracesproul-changed-our-orgs-internal-model-in-fleet-from-so
- claude-sonnet-5-is-now-available-in-cursor-on-cursorbench-it
- glm-51-by-zaiorg-is-now-3-in-code-arena---surpassing-gemini-
- have-been-using-grok-46-last-24-hours-first-takeaway-insanel
- inside-andon-market-billed-as-the-first-retail-boutique-run-
- pair-nova-2-lite-with-claude-for-cost-optimized-document-pro
- switched-out-sonnet-46-for-glm-51-through-fireworksaihq-whil
- swyx-vibhuuuus-aidotengineer-im-ready-for-the-poaster-tomorr
- this-is-big-local-ai-news-a-new-open-source-computer-use-llm
- through-aug-31-sonnet-5-will-use-roughly-30-less-quota-than-
KAN
Mentioned in 13 entries
- a-comparative-analysis-of-mlp-and-kolmogorov-arnold-networks
- aasist3-kan-enhanced-aasist-speech-deepfake-detection-using-
- ar-kan-autoregressive-weight-enhanced-kolmogorov-arnold-netw
- autocorrelation-reintroduces-spectral-bias-in-kans-for-time-
- enhanced-feature-extraction-for-iot-network-intrusion-detect
- enhancing-bigru-with-a-kan-block-for-legal-document-classifi
- flashkan-b-spline-kans-via-truncated-power-form
- fourier-kan-mamba-a-novel-state-space-equation-approach-for-
- kan-mlp-mixer-a-comprehensive-investigation-of-the-usage-of-
- kan-vs-lstm-performance-in-time-series-forecasting
- kan-we-flow-advancing-robotic-manipulation-with-3d-flow-matc
- kolmogorov-arnold-networks-kan-for-aerodynamic-prediction-a-
- mvg-kan-multi-view-geo-wind-guided-kan-for-pm25-forecasting
Gemma4
Mentioned in 13 entries
- 3090-gemma4-qat-mtp-quick-tps-numbers-tldr-12-18x-better
- b11009
- basic-coding-system-demo-via-ollama---quick-head-to-head-gem
- claude-code-gemma4-and-remotion
- day-0-mtp-support-for-gemma4-now-available-at-vllm-with-read
- fixed-ipex-llm-modern-ollama-models-qwen3-gemma4-on-intel-ar
- gemma4-slaying-prompt
- httpsollamacomlibrarygemma4
- httpsollamacomlibrarygemma4tags
- model-page-httpsollamacomlibrarygemma4
- moe-cpu-offload-benchmark-on-deepseek-v4gemma4qwengpt-oss-te
- trying-to-setup-two-llms-to-run-on-2-gpus-separately-but-sim
- we-have-eleven-reader-at-home
YOLOv8
Mentioned in 13 entries
- a-computer-vision-approach-to-visual-fraud-detection-in-phis
- an-exploratory-study-of-frequency-aware-task-weighting-for-y
- attention-augmented-yolov8-with-ghost-convolution-for-real-t
- comparing-yolov11-and-yolov8-for-instance-segmentation-of-oc
- cross-generation-optimization-of-yolov26-yolov11-and-yolov8-
- enhanced-detection-of-tiny-objects-in-aerial-images
- multiscale-real-time-object-detection-in-the-nms-free-era-a-
- performance-analysis-of-yolov11-and-yolov8-for-mixed-traffic
- real-time-automatic-license-plate-recognition-using-yolov8-s
- real-time-industrial-defect-detection-on-edge-hardware-using
- real-time-pedestrian-attribute-recognition-with-yolov8-and-r
- ultralytics-yolo-evolution-an-overview-of-yolo27-yolo26-yolo
- yolov8-to-yolo11-a-comprehensive-architecture-in-depth-compa
Qwen3.8-Flash
Mentioned in 13 entries
- 125b6b-1m-context-multimodal-now-in-opencode-go-happy-coding
- a-high-performance-125b-model-now-running-locally-on-just-75
- alibaba-releases-qwen38-flash-an-open-weight-125b-parameter-
- api-is-live-on-qwencloud-httpswwwqwencloudcommodelsqwen38-fl
- big-thanks-to-sglproject-for-the-day-0-support-qwen38-flash-
- meet-qwen38-flash-a-multimodal-moe-and-an-early-preview-of-t
- qwen37-max-and-qwen38-flash-from-alibabaqwen-are-now-40-off-
- qwen38-flash-is-40-off-through-the-rest-of-the-month-making-
- qwen38-flash-is-now-live-on-openrouter-coding-assistants-age
- qwen38-flash-on-qwencloud-0151m-input-tokens-0471m-output-to
- qwen38-flash-on-rtx3090-64gb-ram-but-you-only-need-12gb-vram
- swift15-qwen38-flash-next-is-phenomenal-vs-base-38-flash
- whats-better-than-an-open-weight-multimodal-model-release-we
Ox Alpha
Mentioned in 13 entries
- brace-yourselves-we-just-entered-a-new-era-of-frontier-multi
- first-serious-confirmation-ox-alpha-is-glm-53-flash
- glm-53-and-glm-53-flash-previously-ox-alpha-are-fully-rolled
- glm-53-flash-aka-ox-alpha-is-now-the-best-intelligence-per-d
- got-confirmation-on-what-the-ox-alpha-model-is-cant-disclose
- i-benchmarked-ox-alpha-on-swe-bench-verified-mini-50-tasks-9
- new-stealth-model-ox-alpha-look-like-glm-or-mimo-new-model
- nobody-knows-who-built-ai-coding-model-ox-alpha-or-where-the
- now-we-know-the-popular-ox-alpha-llm-was-glm-53-flash-compar
- ox-alpha-a-stealth-model-from-an-unknown-ai-lab-with-a-1m-to
- recommended-agent-skill-eli5-works-great-for-visualizing-dee
- the-mystery-model-ox-alpha-is-now-available-in-autoclawaier-
- zai-open-sources-ox-alpha-model-as-glm-53-flash
Muse Spark 1.1
Mentioned in 13 entries
- introducing-muse-spark-11
- many-people-were-doubting-metas-position-in-the-ai-race-yest
- mark-zuckerberg-hasnt-posted-since-2023-x-is-the-global-town
- meta-launches-flagship-muse-spark-11-model-with-multi-agent-
- meta-makes-muse-spark-11-available-to-consumers-debuts-new-f
- meta-prices-muse-spark-11-at-1251m-input-tokens-and-4251m-ou
- meta-releases-muse-spark-11-capable-of-more-advanced-coding-
- meta-updates-meta-ai-with-muse-spark-11-powered-agentic-capa
- metas-muse-spark-11-hacked-an-external-organization-during-c
- muse-spark-11-available-in-new-meta-model-api-somewhere-near
- muse-spark-11-is-now-available-on-ai-gateway
- new-meta-model
- which-of-gpt-56-grok-45-fable-5-or-muse-spark-11-is-least-po
GLM 5.3
Mentioned in 13 entries
- glm-53-and-glm-53-flash-previously-ox-alpha-are-fully-rolled
- glm-53-and-glm-53-flash-ran-locally-on-rtx-pro-6000-ws-and-b
- glm-53-is-now-available-for-training-on-the-serverless-train
- glm-53-is-now-available-in-bionic-agent-us-hosted-with-zero-
- glm-53-is-now-available-in-perplexity-computer-built-for-lon
- glm-53-slopcodebench-results
- glm-53-vs-deepseek-v4-usage
- how-is-anyone-getting-anything-done-with-ollama-cloud
- i-built-a-server-with-768gb-vram-for-frontier-but-all-new-fr
- nice-stats-from-openrouter-that-illustrate-how-togethercompu
- this-model-is-insane-at-3d-website-design-i-asked-glm-53-lef
- try-glm-53
- unpopular-opinion-qwen-38-27b-is-not-an-overthinker
Nemotron 3 Nano Omni
Mentioned in 13 entries
- excited-to-support-nvidia-nemotron-3-nano-omni-now-available
- introducing-nvidia-nemotron-3-nano-omni-long-context-multimo
- nemotron-3-nano-omni-efficient-and-open-multimodal-intellige
- nemotron-3-nano-omni-is-available-locally-on-ollama-this-req
- nemotron-3-nano-omni-is-now-in-lm-studio-a-new-30b-multi-mod
- nvidia-introduces-nemotron-3-nano-omni-with-vision-and-speec
- nvidia-launches-nemotron-3-nano-omni-an-open-multimodal-mode
- nvidia-launches-nemotron-3-nano-omni-model-unifying-vision-a
- nvidia-nemotron-3-nano-omni-is-now-on-together-ai-enterprise
- nvidia-nemotron-3-nano-omni-now-deployed-on-vultr
- nvidia-nemotron-3-ultra-now-available-on-amazon-sagemaker-ju
- some-awesome-open-models-trending-on-hf-right-now-including-
- together-ai-brings-nvidia-nemotron-3-nano-omni-to-developers
Gemini Spark
Mentioned in 13 entries
- anyone-understand-what-google-mean-by-gemini-spark-runs-on-g
- gemini-spark-is-announced-and-will-go-live-next-week-its-goo
- gemini-spark-is-the-most-impressive-and-terrifying-ai-experi
- geminis-new-ai-agent-is-about-as-good-as-googles-demo
- google-io-gemini-spark-antigravity
- google-starts-rolling-out-access-to-gemini-spark-for-google-
- googles-ai-future-demands-trust-and-your-personal-data
- googles-new-gemini-spark-is-an-always-on-ai-agent-for-daily-
- hands-on-with-gemini-spark-beta-rolling-out-to-ai-ultra-subs
- i-dont-have-much-to-say-about-this-years-google-io-because-i
- io-2026-welcome-to-the-agentic-gemini-era
- new-upgrades-to-the-geminiapp-are-you-helping-you-get-more-d
- our-most-intelligent-workhorse-model-yet-for-coding-and-agen
Qwen 3.8 Flash
Mentioned in 13 entries
- 10-performance-improvement-on-moe-ssd-streaming-with-expert-
- 4xradeon-ai-pro-r9700-people-how-are-your-benchmarks
- another-harness-matters-post-codex-cli-pi-and-opencode
- b10712
- computer-use-powered-by-localcloud-models-for-regulated-indu
- deepseek-v4-0731---qwen-38-flash---glm-53-flash-and-back-aga
- learningrsi-through-ngrams
- my-qwen38-flash-next-recipe-for-single-gb10dgx-spark-uses-in
- qwen-38-27b-vs-qwen-38-flash-next-and-time-to-complete-a-cod
- qwen-38-flash-next-can-build-funny-games
- qwen-38-flash-next-q4km-130k-context-q8-cache-on-16gb-vram-a
- tqwen-38-flash-next-ud1s-on-6gb-vram-and-16-gb-system-ram
- what-are-your-experiences-with-using-a-hybrid-cloudlocal-set
GPT-6 Luna
Mentioned in 13 entries
- big-model-release-today---i-wrote-about-claude-opus-55-gpt-6
- bring-more-intelligence-to-everyday-work-with-gpt-6-sol-and-
- gpt-6-luna-is-half-the-price-of-56-luna-which-was-already-an
- gpt-6-sol-and-luna-are-great-models-but-also-these-character
- gpt-6-sol-and-luna-roll-out-today-in-chatgpt-work-and-codex-
- gpt-6-sol-costs-21m-input-and-101m-output-tokens-and-gpt-6-l
- gpt6-sol-and-luna-build-on-astras-advances-in-alignment-show
- holy-fuck-its-christmas
- in-our-evaluations-we-see-that-gpt-6-sol-reads-about-17-less
- interestingly-just-overheard-in-swyx-podcast-interviewing-an
- openai-launches-gpt-6-sol-and-luna-saying-sol-makes-about-ha
- please-welcome-gpt-6-sol-and-gpt-6-luna-to-the-gpt-6-univers
- we-comprehensively-evaluated-16-recent-frontier-vlms---inclu
GLM-5
Mentioned in 12 entries
- breaking-glm-53-flash-by-zaiorg-debuts-at-6th-overall-on-des
- check-out-the-glm-51-first-impressions-with-peter-on-our-you
- come-check-out-glm-51-in-code-arena-for-agentic-web-developm
- glm-5-serving-parameter-tuning-for-openclaw-single-deploymen
- glm-51-by-zaiorg-just-launched-in-the-text-arena-and-is-now-
- glm-51-gives-teams-a-stronger-model-for-coding-tool-use-and-
- glm-51-towards-long-horizon-tasks
- highlights-28-coding-improvement-over-glm-5-with-refined-rl-
- if-youve-encountered-garbled-output-like-this-while-using-gl
- lets-deep-dive-into-glm-model-improvement-by-zaiorg-over-the
- so-excited-for-this---in-the-past-four-months-ive-written-al
- we-partnered-with-zaiorg-to-bring-glm-51-to-modal-free-to-tr
AlphaFold
Mentioned in 12 entries
- a-second-nobel-prize-for-alphafold-alphafold-deepmind-nobelp
- alphafolds-bayesian-roots-in-probability-kinematics
- confornets-latents-based-conformational-control-in-openfold3
- eladgil-bs-attention-was-born-in-montral-pytorch-in-nyc-alph
- every-time-i-tell-ai-utopianists-that-biology-is-too-complex
- finally-a-big-name-has-the-courage-to-tell-it-we-are-nowhere
- he-won-a-nobel-here-for-alphafold-then-he-left---john-jumper
- how-to-accelerate-protein-structure-prediction-at-proteome-s
- i-want-ai-to-succeed-and-have-been-positive-about-many-syste
- sources-google-deepmind-has-reassigned-the-majority-of-the-o
- this-is-an-interesting-point-in-the-new-ted-chiang-piece-no-
- very-interesting-paper-shows-some-suggestive-evidence-that-r
SAM 3
Mentioned in 12 entries
- activesam-fast-and-accurate-open-vocabulary-semantic-segment
- dense-structural-priors-for-sparse-functional-landmark-local
- eovsam-efficient-open-vocabulary-segmentation-with-sam-3-in-
- farmland-extent-and-visible-boundary-mapping-from-1-m-naip-i
- hyperdam-hyperspectral-distractor-aware-memory-with-amodal-e
- lightweight-distillation-of-sam-3-and-dinov3-for-edge-deploy
- omniovcd-streamlining-open-vocabulary-change-detection-with-
- opted-open-preprocessed-trachoma-eye-dataset-using-zero-shot
- prompt-calibrated-sam-3-for-open-vocabulary-remote-sensing-s
- sam3cpp---metas-sam-3-in-pure-c-with-ggerganovs-ggml---suppo
- segearth-ov3-exploring-sam-3-for-open-vocabulary-semantic-se
- single-shot-generation-still-surfaces-net-new-kernels-with-n
GNN
Mentioned in 12 entries
- a-unifying-relational-perspective-on-expressive-lottery-tick
- badimplant-injection-based-multi-targeted-graph-backdoor-att
- enhanced-feature-extraction-for-iot-network-intrusion-detect
- examining-the-efficacy-of-graph-neural-network-message-passi
- gpevac-gnn-based-ppo-for-adaptive-evacuation-routing-during-
- kolmogorov-arnold-networks-kan-for-aerodynamic-prediction-a-
- neighbourhood-transformer-switchable-attention-for-monophily
- on-the-role-of-dag-topology-in-energy-aware-cloud-scheduling
- poisoning-the-inner-prediction-logic-of-graph-neural-network
- rgnn-model-for-fraud-detection-isnt-performing-wellr
- trine-a-token-aware-runtime-adaptive-fpga-inference-engine-f
- warm-starting-pde-solvers-with-any-dimensional-machine-learn
Pi
Mentioned in 12 entries
- harness-showdown-claude-code-vs-opencode-vs-pi-with-deepseek
- highly-recommended-read-aligns-with-what-i-see-in-my-own-har
- pi-agent-users---nvidia-released-sol-pi---a-pi-extension-bas
- pi-huggingface-llamacpp-qwen36
- pi-qwen36-is-like-fsd-12-we-kind-of-have-working-self-drivin
- recommended-agent-skill-eli5-works-great-for-visualizing-dee
- recursive-language-models-claude-code-agent-swarms-big-resea
- run-ornith-with-ollama-ollama-run-ornith-for-coding-use-it-w
- software-engineers-do-you-honestly-get-anything-useful-out-o
- someone-used-pi-to-build-an-excel-agent-then-published-all-t
- switched-from-opencode-to-pi---what-settingsplugins-would-yo
- want-to-test-how-good-glm-53-flash-is-i-have-been-finding-it
Variational Autoencoder
Mentioned in 12 entries
- a-generative-deep-learning-workflow-for-inverse-molecular-de
- a-multi-stage-constrained-optimization-framework-for-data-dr
- domain-specific-latent-representations-improve-the-fidelity-
- elvae-evidential-learning-based-variational-autoencoder-for-
- exact-prior-variational-autoencoder-x-vae-learning-data-adap
- from-baseline-to-follow-up-counterfactual-spine-dxa-image-sy
- heavy-tailed-class-conditional-priors-for-long-tailed-genera
- mitigating-the-reconstruction-detection-trade-off-in-vae-bas
- theoretical-study-on-the-evidential-learning-based-variation
- variational-autoencoder-domain-adaptation-for-cross-system-g
- variational-autoencoder-layer
- variational-objectives-for-amortized-bayesian-inference-in-i
Large Reasoning Models
Mentioned in 12 entries
- accurate-legal-reasoning-at-scale-neuro-symbolic-offloading-
- cut-your-losses-learning-to-prune-paths-early-for-efficient-
- effort-as-ceiling-not-dial-reasoning-budget-does-not-modulat
- enhancing-ai-interpretability-and-safety-through-localised-a
- even-right-now-many-workflows-are-morphing-into-lrm-guided-h
- exploring-the-system-1-thinking-capability-of-large-reasonin
- hauntattack-when-attack-follows-reasoning-as-a-shadow
- reasoning-models-will-sometimes-lie-about-their-reasoning
- restoring-exploration-after-post-training-latent-exploration
- rom-real-time-overthinking-mitigation-via-streaming-detectio
- self-reset-learning-to-self-recover-from-unsafe-reasoning-tr
- when-models-outthink-their-safety-unveiling-and-mitigating-s
diffusion models
Mentioned in 12 entries
- climate-physics-dynamic-matching
- controllable-blind-deblurring-with-diffusion-models
- dgs-net-distillation-guided-gradient-surgery-for-clip-fine-t
- fundamental-limitation-in-explaining-ai
- icg-improving-cover-image-generation-via-mllm-based-promptin
- lhsdet-high-resolution-ai-generated-image-detection-via-visu
- muflow-leveraging-average-images-for-improving-generalisatio
- ns-net-decoupling-clip-semantic-information-through-null-spa
- performance-optimization-and-comparative-analysis-of-generat
- seed-to-seed-unpaired-image-translation-in-diffusion-seed-sp
- towards-high-fidelity-face-swapping-a-comprehensive-survey-a
- variational-grey-box-dynamics-matching
Mistral
Mentioned in 12 entries
- android-app-ollama-talk-now-on-playstore
- austria-is-rolling-out-a-government-ai-platform-using-mistra
- i-gave-five-different-local-llms-a-town-they-invented-facebo
- i-ran-a-prompt-injection-test-suite-against-qwen25-7b14b-and
- making-knowledge-accessible-divergent-readability-accuracy-s
- ollama-mistral-with-mcp-to-mempalace
- ollama-pull-mistral---ai-only-explains-the-process-and-wont-
- pruning-unsafe-tickets-a-resource-efficient-framework-for-sa
- research-entity-extraction-and-topic-detection-from-ukri-gra
- the-washington-post-processed-179b-input-tokens-per-month-th
- there-are-ai-applications-that-are-equally-inside-the-fronti
- weve-got-all-the-models-here-httpsdellhuggingfacecoauthentic
Muse Glimmer 30B
Mentioned in 12 entries
- 12gb-vram-gang-whats-our-plan
- achievable-253-ts---unslothmuse-glimmer-30b-ud-q5km-on-a-509
- dont-sleep-on-exl3-quants
- ds4-cloud-30-min-vs-qwen36-36b-2-min-vs-muse-glimmer-30b-3-m
- got-metas-new-muse-glimmer-30b-running-on-my-macbook-m3-max-
- interesting-uses-for-muse-glimmer-30b
- local-benchmark-muse-glimmer-30b-vs-qwen-36-27b-vs-gemma4-31
- meta-released-muse-glimmer-30b-multimodal-model-for-your-cla
- muse-glimmer-30b-is-now-available-on-fireworks-dedicated-tra
- muse-glimmer-the-new-30b-model-is-available-on-hugging-face-
- qwen36-35b-2-min-vs-muse-glimmer-30b-4-min-on-custom-llamacp
- you-can-now-try-offline-computer-use-with-cua-driver-muse-gl
Nemotron 3.5 Lightning
Mentioned in 12 entries
- dense-vs-moe-models-active-parameters-throughput-and-when-to
- developing-nemotron-35-lightning-nvfp4-with-qad-using-nvidia
- frontier-reasoning-reaches-the-edge-how-to-deploy-and-optimi
- looking-for-a-faster-specialized-model-for-your-agent-work-n
- nemotron-35-lightning-is-available-in-lm-studio-the-model-is
- nemotron-35-lightning-is-available-on-together-ai-through-de
- nvidia-nemotron-35-lighting-is-available-on-ollama-its-a-30b
- nvidia-nemotron-35-lightning-and-nemo-switchyard-deliver-fas
- nvidia-nemotron-35-lightning-is-now-live-on-together-ai-the-
- nvidia-releases-nemotron-35-lightning-an-open-30b-parameter-
- nvidia-releases-nemotron-35-lightning-and-nemo-switchyard-to
- tested-nemotron-35-lightning-locally-on-coding-hermes-agent-
GPT-3
Mentioned in 12 entries
- ai-alarm-hype-cycles---a-brief-history-gpt-2-openai-withhold
- blast-from-the-past-35-years-ago-some-things-have-changed-es
- four-years-later-i-got-the-new-claude-fable-5-to-actually-bu
- from-gpt-3-to-gpt-5-mapping-their-capabilities-scope-limitat
- gpt-3-is-discontinued-today
- gpt-6-is-pretty-good-tbh
- interesting-to-listen-to-this-years-later-what-held-up-what-
- llm-utility-has-increased-and-deeper-understanding-to-a-degr
- one-shotting-a-raccoon-heist-game-using-claude-fable-5
- remind-me-elonmusk-was-gpt-5-really-smarter-than-the-smartes
- seems-obvious-now-but-4-years-ago-i-was-getting-blank-stares
- wheres-the-raccoon-with-the-ham-radio-chatgpt-images-20
Gemini 3.8 Flash
Mentioned in 12 entries
- check-out-this-weeks-shipping-recap-gemini-38-flash-our-most
- dont-count-out-gemini-just-yet-feels-like-a-new-era-of-model
- gemini-38-flash-might-be-maxxed-on-other-benchmarks-but-not-
- google-launches-gemini-38-flash-cyber-for-partners-in-its-ne
- google-launches-two-gemini-38-models-with-cutting-edge-reaso
- introducing-gemini-38-flash-another-jump-in-geminis-agentic-
- lwiai-podcast-256---fable-51-astra-tease-gemini-38-flash
- rolling-out-starting-today-developers-both-models-in-googlea
- sources-google-plans-to-release-gemini-38-flash-as-soon-as-w
- update-yandexaliceai-80b-a3b-fine-tune-progress
- were-introducing-gemini-38-flash-built-to-tackle-complex-age
- weve-benchmarked-all-the-latest-frontier-models-on-hard-docu
DeepSeek-V4-Flash-Vision-Exp
Mentioned in 12 entries
- b10762
- b11049
- b11075
- b11081
- b11172
- b11183
- brace-yourselves-we-just-entered-a-new-era-of-frontier-multi
- deepseek-harness-v011-released
- deepseek-v4-flash-vision-exp-285b-moe-on-10-12x-rtx-3090-spe
- deepseek-v4-flash-vision-exp-is-amazing-at-creating-game-wor
- deepseek-v4-flash-vision-exp-is-now-live-on-the-deepseek-api
- ram-offloading-with-vllm---tcclaviger-appreciation-post
Gemini Omni Flash
Mentioned in 12 entries
- bringing-speed-and-strong-cost-performance-to-the-market-wit
- gemini-omni-flash-can-manipulate-objects-and-environments-in
- gemini-omni-flash-can-swap-objects-change-environments-and-m
- gemini-omni-flash-image-to-video-httpslinkscomfyorg4y3e93o
- gemini-omni-flash-is-now-in-comfyui-via-partner-nodes-google
- gemini-omni-flash-video-edit-httpslinkscomfyorg445s9am
- google-ai-plus-pro-and-ultra-subscribers-can-gemini-omni-fla
- google-launches-nano-banana-2-lite-a-low-cost-text-to-image-
- googles-gemini-omni-flash-and-nano-banana-2-lite-support-sli
- httpsxcomraza8542121status2056807415421751372
- start-building-with-nano-banana-2-lite-and-gemini-omni-flash
- we-were-able-to-sit-down-with-the-googledeepmind-team-behind
Gemini Enterprise
Mentioned in 12 entries
- day-2-at-google-cloud-next-a-marathon-developer-keynote
- developers-guide-to-gemini-enterprise-and-a2ui-integration
- google-cloud-partners-deliver-new-security-agents-and-ai-def
- google-cloud-says-its-spacex-compute-deal-is-a-short-term-ag
- google-puts-gemini-enterprise-at-the-heart-of-the-new-agenti
- lookers-semantic-layer-governs-gemini-enterprise-data-for-us
- now-introducing-gemini-enterprise-for-financial-services
- now-introducing-gemini-enterprise-for-legal
- pythian-gemini-enterprise-database-operations
- reimagining-work-how-pythians-internal-ai-playbook-delivers-
- small-and-midsize-businesses-jumpstart-their-ai-transformati
- socradar-powers-rapid-threat-detection-with-alloydb-and-gemi
DeepSeek-V4.1-Flash
Mentioned in 12 entries
- big-overhaul-on-deepseek-v41-using-an-encoder-decoder-setup-
- deepseek-debuts-deepseek-v41-flash-its-smallest-model-built-
- deepseek-releases-v41-flash-says-it-outperforms-flagship-v4-
- deepseek-v41-flash-is-now-being-rolled-out-on-ollamas-cloud-
- deepseek-v41-flash-is-now-fully-rolled-out-and-available-on-
- deepseek-v41-flash-is-now-rolling-out-for-pro-plan-subscribe
- deepseek-v41-flash-pushing-the-limits-of-kv-cache-compressio
- hoping-for-optimized-smarter-upcoming-models-like-deepseek-v
- i-enjoyed-the-daily-hf-papers-today
- introducing-deepseek-v41-flash-smarter-faster-more-efficient
- model-grafting-turning-qwen35-4b-into-a-causal-encoder-decod
- now-live-in-lm-studio-bionic
Qwen-Image-2.1
Mentioned in 12 entries
- day-0-openvino-support-from-inteldevs-qwen-image-21-is-ready
- i-tested-qwen-image-21-its-amazing
- meet-qwen-image-21-the-most-balanced-and-cost-effective-imag
- official-qwen-image-21-blog-post-detailing-its-capabilities-
- qwen-image-21-also-improves-visual-quality-over-its-predeces
- qwen-image-21-huggingapps-live-demo-on-spaces-one-single-che
- qwen-image-21-is-now-supported-in-comfyui-try-it-now-and-sha
- qwen-image-21-released
- qwen-image-21-supports-a-variety-of-editing-tasks-while-bala
- thanks-arena-for-the-recognition-qwen-image-21-is-now-the-1-
- thanks-sglproject-for-the-day-0-support-sglang-diffusion-now
- thanks-vllmproject-for-the-day-0-support-generate-edit-trans
GAN
Mentioned in 11 entries
- a-gan-and-llm-driven-data-augmentation-framework-for-dynamic
- an-improved-generative-adversarial-network-for-micro-resisti
- assessment-of-cloud-and-associated-radiation-fields-from-a-g
- explainable-ai-for-mental-health-prediction-in-drug-affected
- fast-fourier-convolutional-gan-for-30-m-clear-sky-land-surfa
- i-built-ganfs-a-python-package-that-uses-gans-to-automate-fe
- inverse-design-of-realizable-metasurface-based-absorbers-usi
- mmwave-radar-aware-dual-conditioned-gan-for-speech-reconstru
- nepscript-genesis-neural-architecture-search-for-handwritten
- pcp-gan-property-constrained-pore-scale-image-reconstruction
- swin-meets-efficientnet-lightweight-architectures-for-gan-ba
DeepSeek-R1
Mentioned in 11 entries
- asking-like-socrates-socrates-helps-vlms-understand-remote-s
- knowledge-distillation-from-large-reasoning-models-to-compac
- lead-length-efficient-adaptive-and-dynamic-reasoning-for-lar
- lisa-linear-indexed-sparse-attention-for-efficient-long-cont
- longflow-efficient-kv-cache-compression-for-reasoning-models
- more-thinking-more-bias-length-driven-position-bias-in-reaso
- reasoning-models-can-be-accurately-pruned-via-chain-of-thoug
- smartthinker-progressive-chain-of-thought-length-calibration
- solving-physics-olympiad-via-reinforcement-learning-on-physi
- wait-wait-wait-why-do-reasoning-models-loop
- when-to-retrieve-during-reasoning-adaptive-retrieval-for-lar
Kimi K2.5
Mentioned in 11 entries
- bro-wtf-qwen-lab-cooked-with-qwen-38-27b-its-so-fucking-good
- composer-25-is-built-on-the-same-open-source-base-as-compose
- cursor-releases-composer-25-saying-its-better-at-sustained-w
- kimi-k25-visual-agentic-intelligence
- meet-kimi-k26-agent-swarm-highlights-swarms-elevated---300-p
- silicon-valley-is-quietly-running-on-chinese-open-source-ai-
- the-cursorai-team-shipped-composer-2-and-now-composer-25-on-
- tried-ollama-cloud-just-realize-only-kimi-model-accept-image
- were-launching-kimi-k26-on-fireworks-as-a-day-0-launch-partn
- were-the-1-open-model-in-vision-and-document-arena
- weve-got-all-the-models-here-httpsdellhuggingfacecoauthentic
nnU-Net
Mentioned in 11 entries
- advanced-tumor-segmentation-in-petct-imaging-a-training-stra
- assessing-nnu-net-generalization-across-brain-tumor-populati
- automated-lesion-segmentation-of-stroke-mri-using-nnu-net-a-
- budget-aware-uncertainty-for-radiotherapy-segmentation-qa-us
- detail-consistent-stage-wise-distillation-for-efficient-3d-m
- lesion-segmentation-in-moderate-to-severe-traumatic-brain-in
- lightmedseg-isles-stroke-lesion-segmentation-with-81x-fewer-
- nnu-net-scaling-nnu-net-with-prefix-based-data-augmentation
- pre--and-post-treatment-brain-metastases-segmentation-using-
- strokeseg2-stroke-lesion-segmentation-in-clinical-research-w
- topology-driven-fusion-of-nnu-net-and-mednext-for-accurate-b
Qwen3.6 27B
Mentioned in 11 entries
- claude-code-opus-48-vs-local-qwen36-27b-one-shot-coding-benc
- i-dont-think-i-need-cloud-models-anymore
- pi-gives-you-wings-wherever-you-are
- qwen36-27b-35b-on-vllm-single-r9700-gfx1201
- qwen36-27b-is-now-in-lm-studio-vision-reasoning-and-agentic-
- qwen36-27b-on-dual-rtx-5060-ti-16gb-with-vllm-60-toks-204k-c
- tesla-v100-qwen36-27b-performance
- tested-in-coding-q8kxl-qwen38-27b-vs-bf16-qwen36-27b
- these-are-some-of-my-llm-assisted-contributions-from-the-pas
- this-is-where-we-are-right-now-and-im-not-gonna-lie-it-feels
- today-were-announcing-bonsai-27b-the-first-27b-class-model-t
Qwen3.7-Max
Mentioned in 11 entries
- alibaba-releases-qwen37-plus-a-multimodal-proprietary-model-
- cowork-productivity-assistantqwen37-max-serves-as-your-advan
- for-anyone-who-isnt-sure-this-is-how-you-release-a-model-and
- implicit-caching-is-now-live-on-qwen37-max-kicks-in-automati
- introducing-qwen37-max-from-alibabaqwen-qwens-flagship-model
- meet-qwen37-max-our-latest-flagship-made-for-the-agent-era-a
- qwen37-max-and-qwen38-flash-from-alibabaqwen-are-now-40-off-
- qwen37-max-just-hit-3-on-itbench-aa-a-fresh-benchmark-testin
- qwen37-max-just-hit-4-on-code-arena-on-par-with-claude-opus-
- qwen37-max-just-landed-at-566-on-the-artificial-analysis-int
- try-qwen37-max-now-on-together-ai-httpwwwtogetheraimodelsqwe
DiffusionGemma
Mentioned in 11 entries
- diffusiongemma-4x-faster-text-generation
- diffusiongemma-technical-report
- google-deepmind-releases-diffusiongemma-a-model-that-runs-lo
- google-introduces-diffusiongemma-an-experimental-26b-paramet
- google-open-sources-speedy-diffusiongemma-text-diffusion-mod
- how-transparent-is-diffusiongemma
- jev-style-structured-decisions-with-diffusiongemma-gguf-in-t
- meet-diffusiongemma-an-experimental-open-model-that-explores
- nvidia-accelerates-google-deepminds-diffusiongemma-for-local
- oui-1-a-model-that-generates-bespoke-ui-elements
- run-diffusiongemma-on-nvidia-for-developer-ready-high-throug
Fugu
Mentioned in 11 entries
- fugu-stands-shoulder-to-shoulder-with-leading-models-like-fa
- introducing-tinyrouter-i-reverse-engineered-the-routing-arch
- model-orchestration-is-in-many-ways-the-natural-outgrowth-of
- product-designer-httpssakanaaicareersproduct-designer-sakana
- sakana-aigpt-6-astrafable-51fugu-httpswwwitmediacojpaiplusar
- sakana-chat-just-got-a-big-upgrade-no-login-required-free-to
- sakana-fugu-technical-report-httpsgithubcomsakanaaifugublobm
- sakanaailabs-sign-up-below-to-get-a-bunch-of-free-api-credit
- tokyo-based-sakana-ais-fugu-and-china-based-360s-cybersecuri
- we-just-pushed-a-big-update-to-sakana-chat-no-login-required
- were-excited-to-collaborate-with-nvidia-to-build-the-next-ge
DeepONet
Mentioned in 11 entries
- a-deeponet-for-inverting-the-neumann-to-dirichlet-operator-i
- a-trainable-by-parts-operator-learning-framework-bridging-de
- conformalized-quantum-deeponet-ensembles-for-scalable-operat
- dvl-deeponet-a-physics-guided-operator-learning-for-resilien
- fast-deeponet-factor-augmented-branch-representations-for-hi
- fpinn-deeponet-a-physics-informed-operator-learning-framewor
- learning-the-helmholtz-equation-operator-with-deeponet-for-n
- multi-head-residual-gated-deeponet-for-coherent-nonlinear-wa
- quantum-sedonet-spectrally-embedded-quantum-deep-operator-ne
- rationally-enriched-chebyshev-trunk-bases-for-deeponet-surro
- spectral-embedding-via-chebyshev-bases-for-robust-deeponet-a
Ling-3.0-flash
Mentioned in 11 entries
- antlingve-open-sourced-6-base-model-checkpoints-for-ling-30-
- higher-acceptance-length-slower-prose-lings-n123-mtp-test-on
- inclusionailing-30-flash-hugging-face
- inclusionailing-30-flash-vl-hugging-face
- inclusionailing-30-flash-weights-are-up-on-hugging-face-mit-
- inclusionailing-30-tiny-8b-a13b-moe-hugging-face
- ling-30-flash-is-another-potential-model-to-test-before-qwen
- ling-30-flash-the-new-moe-model-from-antlingagi-is-now-free-
- ling-30-flash-vl-built-on-ling-30-flash-with-visual-understa
- ling-30-flash-weights-sglang-says-day-0-vllm-says-when-they-
- model-add-ling-30-vl-support-by-aetherbird-pull-request-2915
Qwen3.8-27b
Mentioned in 11 entries
- androidlife-can-an-ai-agent-survive-a-day-in-the-life-of-a-r
- another-qwen38-27b-appreciation-post
- cybertiel-35b-a3bs-uncensored-4-bit-quant-beats-opus-46-medi
- i-did-it-im-free-its-been-7-hours-since-i-used-claudecode
- new-benchmark-just-dropped
- qwen38-27b-has-the-highest-level-of-agency-ive-ever-seen-in-
- qwen38-27b-is-the-first-local-model-im-able-to-blindly-trust
- qwen38-27b-on-rtx-3090---82-tps-single-request-up-to-672-tps
- qwen38-27b-q8-kv-cache-does-seem-to-actually-hurt-model-perf
- qwen38-flash-next-on-509064gb-with-llamacpp---seems-to-not-u
- tunedabliterated-qwen38-27b-into-a-24gb-card-262k-guff-using
PPO
Mentioned in 11 entries
- backprop-free-pong-pc-distributional-hebbian-plasticity-vs-p
- backpropagating-through-simulation-analytic-policy-gradients
- browser-demo-of-our-clash-royale-rl-environment-a-56k-parame
- clashroyaleai-an-open-source-deterministic-clash-royale-simu
- easyppo-stabilizing-the-critic-is-key
- gpevac-gnn-based-ppo-for-adaptive-evacuation-routing-during-
- hyperparameters-fine-tuning-for-marl-comparative-study-d
- loopermuscle-fast-and-stable-learning-of-humanoid-whole-body
- progressive-generalization-augmentation-with-deeply-coupled-
- rtk-vision-ppo-for-autonomous-micro-uav-recovery-on-an-airbo
- sla-safe-energy-control-for-ai-native-ng-ran-using-stability
GPT-5.5 Instant
Mentioned in 11 entries
- gpt-55-instant-is-rolling-out-now-in-chatgpt
- gpt-55-instant-is-rolling-out-over-the-next-two-days-as-the-
- gpt-55-instant-is-starting-to-roll-out-in-chatgpt
- gpt-55-instant-is-starting-to-roll-out-in-chatgpt-its-a-big-
- gpt-55-instant-smarter-clearer-and-more-personalized
- gpt-55-instant-system-card
- openai-claims-chatgpts-new-default-model-hallucinates-way-le
- openais-gpt-55-instant-the-good-the-bad-and-the-insane
- openais-new-gpt-55-instant-makes-chatgpt-smarter-with-more-c
- we-have-a-new-version-of-gpt-55-instant-for-you-and-its-much
- we-shipped-gpt-55-instant-today-to-chat-its-rolling-out-over
ChatGPT Work
Mentioned in 11 entries
- anyone-know-if-chatgpt-codex-in-the-new-chatgpt-desktop-app-
- chatgpt-images-25-is-rolling-out-today-to-all-chatgpt-chatgp
- generating-running-routes-with-gpt-6-astra-and-chatgpt-work
- i-am-so-confused-by-chatgpt-v-chatgpt-codex-v-chatgpt-work-v
- now-in-preview-the-chatgpt-desktop-app-for-linux-use-chatgpt
- openai-debuts-chatgpt-work-an-agentic-tool-for-automating-bu
- openai-launches-a-chatgpt-desktop-app-for-linux-in-preview-s
- openai-rolls-out-gpt-56-after-government-greenlight-and-anno
- openai-unveils-chatgpt-for-financial-services-a-version-of-c
- the-sun-is-out-today
- this-is-pretty-neat-chatgpt-work-and-gpt-6-astra-i-used-max-
Terra
Mentioned in 11 entries
- accessing-openai-models-on-amazon-bedrock-from-australia-wit
- good-new-first-sol-is-a-smart-efficient-and-a-significant-st
- gpt-56-sol-along-with-terra-and-luna-will-launch-publicly-th
- gpt-56-sol-terra-and-luna-are-now-available-in-cursor-on-cur
- i-like-choices-but-now-i-have-2x-modes-codex-vs-work-mode-3x
- introducing-cross-region-inference-for-openai-gpt-56-models-
- introducing-openai-models-on-amazon-bedrock-for-in-country-i
- openai-releases-three-versions-of-gpt-56-called-sol-terra-an
- openai-says-gpt-56-sol-along-with-terra-and-luna-will-launch
- sol-terra-and-luna-our-gpt56-family-of-models-are-starting-t
- the-new-gpt-56-family-luna-terra-sol
Deepseek V4 Flash
Mentioned in 11 entries
- deepseek-v4-flash---hy3-or-is-qwen36-27b-still-the-most-soli
- deepseek-v4-flash-2-bit-quant-is-the-first-model-i-can-run-l
- deepseek-v4-flash-is-now-2-open-weight-model-to-kimi-k3-and-
- deepseek-v4-flash-is-now-free-via-nous-portal-for-a-limited-
- deepseek-v4-flash-is-one-of-the-best-worker-models-out-there
- deepseek-v4-flash-just-hit-colibri-does-anyone-have-numbers
- deepseek-v4-flash-on-slopcodebench
- i-remember-a-time-when-flash-meant-32b
- updated-benchmark-deepseek-v4-flash-on-slopcodebench-local
- will-ollama-upgrade-deepseek-v4-flash-on-cloud
- with-release-of-deepseek-v4-i-wanted-see-how-the-model-sizes
DeepSeek v4
Mentioned in 11 entries
- ainews-moonshot-kimi-k26-the-worlds-leading-open-model-refre
- deepseek-v4-just-dropped
- deepseek-v4-the-most-expected-open-source-model-ever-release
- deepseek-v4-works-fine-but-its-not-the-frontier-pressing-mom
- fine-tuning-cactus-needle-2-can-match-deepseek-v4-on-the-spe
- given-gb300-i-would-estimate-this-is-was-trained-on-about-1e
- i-hope-the-upgrade-to-deepseek-v4-will-make-the-bot-comments
- imo-deepseek-v4-demonstrated-utter-confidence-and-competence
- last-week-in-ai-340---openai-vs-musk-microsoft-deepseek-v4-v
- my-first-two-tikz-sparks-unicorns-from-deepseek-v4-expert-mo
- pulled-the-trigger-today-and-switched-100-of-lindy-traffic-t
Grok 4.7
Mentioned in 11 entries
- and-heres-the-grok-47-model-card-a-few-jumps-vs-46-that-stoo
- breaking-grok-47-dominates-the-legal-agent-benchmark-it-scor
- claude-opus-55-gpt-6-sol-gpt-6-luna-and-a-new-price-war
- grok-47-is-now-available-on-amazon-bedrock
- grok-47-places-spacexai-as-third-after-anthropic-openai-for-
- grok-47-rank-1-on-aa-cyber-index
- i-just-tried-grok-47-and-it-is
- interesting-grok-46-releases-around-august-7-this-will-be-th
- mimo-v26-pro-ties-grok-47-xhigh-and-beats-glm-53-max-on-arti
- try-grok-46-grok-47-which-is-a-major-upgrade-is-coming-soon
- we-put-grok-47-from-spacexai-to-work-on-a-2-million-insuranc
LTX-2
Mentioned in 10 entries
- another-continuous-minutes-long-ltx-2-long-video-the-last-st
- anyone-else-using-ltx-locally-on-mac-via-draw-things-heres-a
- culturescore-evaluating-cultural-faithfulness-in-video-gener
- has-everyone-moved-onto-ltx-23-then
- i-released-a-new-ltx-focused-update-for-deno-custom-nodes-fo
- ltx-23---image-audio-video-controlnet-ic-lora-to-video
- omninft-a-lora-that-improves-the-quality-of-ltx-2
- phosphene-local-video-and-audio-generation-for-apple-silicon
- update-video-outpainting-node-updated-with-ltx-2-support
- whats-the-equivalent-of-ltx-2-raw-format-video-transfer-of-w
MLLMs
Mentioned in 10 entries
- forgefine-grained-multimodal-evaluation-for-manufacturing-sc
- keep-it-simple-multi-key-episodic-memory-retrieval-for-ultra
- llavaflow-preserving-latent-alignment-flow-for-parameter-eff
- modeling-complex-behaviors-multi-personality-composition-and
- points-long-adaptive-dual-mode-visual-reasoning-in-mllms
- small-vision-language-models-are-smart-compressors-for-long-
- spatialmosaic-a-multiview-vlm-dataset-for-partial-visibility
- think-when-needed-adaptive-reasoning-driven-multimodal-embed
- ui-agile-advancing-gui-agents-with-effective-reinforcement-l
- vsi-visual-subtitle-integration-for-keyframe-selection-to-en
MiniMax M2.7
Mentioned in 10 entries
- dell-nvidia
- minimax-m27-advances-scalable-agentic-workflows-on-nvidia-pl
- minimax-m27-is-available-on-ollamas-cloud-and-is-licensed-fo
- minimax-m27-is-live-on-ai-gateway
- minimax-m27-is-now-available-in-lm-studio-this-model-excels-
- minimax-m27-is-now-on-together-ai-trained-by-letting-it-run-
- proud-to-power-nousresearchs-hermes-agent-with-minimax-m27-a
- we-give-you-model-choice-without-infrastructure-chaos-michae
- were-delighted-to-announce-that-minimax-m27-is-now-officiall
- were-thrilled-to-announce-minimaxai-m27-is-now-available-day
OpenAI Codex
Mentioned in 10 entries
- a-pelican-for-gpt-55-via-the-semi-official-codex-backdoor-ap
- agents-dont-paginate-first-chunk-selection-for-llm-tool-resp
- ainews-gpt-55-and-openai-codex-superapp
- every-single-ai-app-is-down
- i-think-the-most-underated-part-of-todays-launch-is-not-gpt-
- if-you-have-been-experiencing-issues-with-openai-codex-oauth
- openclaw-and-claude-can-put-your-ai-generated-podcasts-in-sp
- quoting-openai-codex-baseinstructions
- sanity-checks-for-agentic-data-science
- vercel-plugin-now-supported-on-openai-codex-and-codex-cli
Gemma 4 26B
Mentioned in 10 entries
- for-chat-and-qa-which-moe-model-is-better-qwen-36-35b-or-gem
- gemma-4-26b-31b-from-googledeepmind-is-now-available-on-the-
- gemma-4-on-ai-gateway
- gemma26b-thinking-issue-in-openwebui
- help-gemma-4-26b-lora-training-on-16gb-vram-loss-decreases-b
- lately-ive-been-having-fun-with-running-coding-agents-fully-
- learn-how-to-run-a-local-coding-agent-use---pi-agent---gemma
- moe-models-around-a2b
- need-help-fixing-gemma-4-26b-a4b-toolcalls-in-pi
- turbo-fieldfare-open-source-engine-running-gemma-4-26b-in-2-
Flux 2
Mentioned in 10 entries
- anime-2-real-with-flux-2-klein-9b
- flux-2-klein-9b-controlnets
- flux-2-klein-destiled-my-workflow-following-numerous-request
- flux-2flux-2-klein-transparent-background-lora
- how-much-is-the-quality-loss-in-flux-2-klein
- is-there-a-complication-with-zeta-chromas-training
- is-there-a-way-to-fix-the-flux-2-klein-body-horror-that-come
- its-possible-flux-2-klein-write-texto-100-ok-with-prompt
- running-flux-2-dev-on-5070ti-and-5060-ti-16gb
- what-are-the-current-best-models-quality-wise
Stable Diffusion XL
Mentioned in 10 entries
- 1893-mishima-zaibatsu-patent-files---tekken-fan-concept-with
- 2-characters-with-different-emotions-in-sdxlillu
- better-indoor-backgrounds-with-illustrious-checkpoints
- fine-tuned-sdxl-model-with-lora-to-generate-tribal-indian-ar
- full-sdxl-running-on-iphone-fully-on-device-768px-20-steps-2
- illustrious-z
- multimodal-conditioning-of-fine-tuned-stable-diffusion-xl-fo
- sdxl-workflow
- workflow-for-generating-images-with-wai-illustrious-with-ope
- z-image-turbo-finetune-of-absurd-reality
Z-Image
Mentioned in 10 entries
- can-you-use-qwen35-4b-gemma-4-e4b-with-z-imageturbo
- how-can-i-force-z-image-to-create-full-body-portraits
- i-finetuned-qwen3-17b-to-imitate-original-z-image-text-encod
- inpaint-workflows-for-z-image-qwen-and-flux-fill-onereward
- osaka-tennoji-war-ai-music-video-ltx23-z-image-ace-step-25-x
- some-fruit-comparisons-z-image-vs-ideogram4
- z-anime---full-anime-fine-tune-on-z-image-base
- z-image-an-efficient-image-generation-foundation-model-with-
- z-image-is-unbelievably-good-at-anime-prompts-given
- z-image-turbo-checkpoint---deedeemegadoodo-edition
o1
Mentioned in 10 entries
- 55-is-an-autistic-genius-with-very-strange-taste-in-naming-s
- 55-is-so-earnest-little-engine-that-could-energy
- ai-keeps-getting-better-but-the-last-time-the-shape-of-the-j
- goblin-level-blog-post
- its-the-graph-again-not-the-metr-graph-the-one-from-the-o1-l
- new-paper-on-an-old-ai-tests-o1-against-doctors-on-medical-b
- openai-o1-system-card
- study-openais-o1-correctly-diagnosed-67-of-emergency-room-pa
- when-to-retrieve-during-reasoning-pay-attention-to-this-one-
- why-is-there-no-thinker-models-with-tokens-for-entire-senten
Wan2.2
Mentioned in 10 entries
- collaborative-few-step-distillation-and-low-bit-quantization
- face-expression-and-lip-sync-in-wan22-animate-workflow
- is-there-a-new-wan22-lightx2v-20260412-fp8-version-for-comfy
- is-training-a-wan-lora-going-to-fix-the-identity-shift-in-i2
- my-first-longer-wan22-continuation-generation-i-am-so-excite
- tail-aware-hifloat4-w4a4-post-training-quantization-for-wan2
- wan22---tips-for-maximizing-video-quality-balancing-motion-a
- wan22-continues-to-outperform-ltx23
- wan22-loras-and-identity-consistency
- would-a-2nd-hand-custom-built-2080-ti-22gb-vram-be-worth-it-
DALL-E 3
Mentioned in 10 entries
- can-we-replicate-2003-dalle-3-yet
- given-how-good-gpt-image-2-is-its-an-absolute-puzzle-openai-
- gpt-image-2-has-created-this-is-there-any-problem-in-its-pro
- gpt-image-2-prompt-that-is-viral-right-now-redraw-the-attach
- gpt-image-guides-20---2025-vs-2026
- models-that-produce-the-same-kind-of-images-and-style-like-d
- presumably-gpt-imagegen-2-aka-chatgpt-images-20-aka-gpt-imag
- same-prompt-with-dalle-2-vs-dalle-3-vs-gpt-image-2
- text-to-image-models-and-their-representation-of-people-from
- trying-image-gen-2
Klein
Mentioned in 10 entries
- 2-the-same-message-applies-to-several-models-chroma-z-image-
- 3-the-same-message-applies-to-several-models-chroma-z-image-
- 4-a-mesma-mensagem-aplica-se-a-vrios-modelos-chroma-z-image-
- 5-the-same-message-applies-to-several-models-chroma-z-image-
- a-fully-character-driven-fantasy-story-made-entirely-with-lt
- anybody-else-find-klein-image-generation-on-musubi-tuner-or-
- flux-2-klein-destiled-my-workflow-following-numerous-request
- flux2-klein-pipeline-for-real-time-webcam-stream-processing-
- same-prompt-for-various-models---chroma-z-image-klein-qwen-e
- whats-new-for-bfl---fluxklein
Fourier Neural Operators
Mentioned in 10 entries
- adex-fno-a-unified-ambient-domain-framework-for-fourier-neur
- can-we-formally-verify-neural-pde-surrogates-smt-compilation
- generic-fno-embedding-energy-conservation-and-entropy-produc
- isotropic-fourier-neural-operators
- multi-scale-dynamic-wake-modeling-of-floating-offshore-wind-
- neural-adjoint-method-for-meta-optics-accelerating-volumetri
- real-vs-complex-spectral-bases-for-neural-operators-the-role
- selection-representation-and-execution-in-sparse-fourier-neu
- sirenfno-efficient-and-full-frequency-learning-of-fourier-ne
- sle-fno-single-layer-extensions-for-task-agnostic-continual-
WAN 2.2
Mentioned in 10 entries
- a-wan-22-post-training-quant-1-model-instead-of-high-low
- any-way-to-do-10sec-video-with-wan-22-firstlast-frame
- can-you-use-a-normal-14b-lora-with-the-wan-14b-aio-model
- current-best-local-model-to-create-accurate-professional-i2v
- exo2egosyn-unlocking-foundation-video-generation-models-for-
- extending-wan-22-t2v-workflows
- forge-neo-issues-with-wan-22-i2v
- how-do-i-install-wan-22-into-forge-neo
- revisiting-wan-22-for-real-person-realism-consented-lora-ret
- training-wan-22-action-loras-on-16-gb-vram-64-gb-ram-managed
Convolutional Neural Networks
Mentioned in 10 entries
- adapting-visualization-techniques-for-time-series-anomaly-de
- balancing-uncertainty-and-diversity-of-samples-leveraging-di
- can-visual-mamba-improve-ai-generated-image-detection-an-in-
- deep-learning-approach-to-bearing-and-induction-motor-fault-
- deep-learning-based-on-generative-adversarial-and-convolutio
- mitigating-shortcut-learning-texture-penalized-prototype-net
- performance-and-interpretability-of-convolutional-transforme
- state-of-health-estimation-using-convolutional-and-bidirecti
- vision-transformers-and-convolutional-neural-networks-for-la
- visionhope-visual-backbones-as-self-modifying-learning-syste
MiniMax-H3
Mentioned in 10 entries
- can-minimax-h3-reason-about-the-physical-world-an-evaluation
- data-free-flow-self-distillation-for-few-step-video-generati
- h3-world-turning-language-understanding-into-world-control
- minimax-h3-38-gb-less-vram-with-runtime-lora-bypass-dora-dyn
- minimax-h3-now-on-huggingface
- pipenetworkminimax-h3-mlx
- qwen-38-max-and-minimax-h3-within-hours-of-each-other
- sol-h3-recursive-self-improvement-for-minimax-h3-inference-a
- spongebob-minimax-h3-test-4-x-5-seconds-turbo-6-steps-134476
- visionary-krea-2-minimax-h3-inference-and-lora-training-in-o
Qwen3.5-4B
Mentioned in 10 entries
- agents-a1-4b-qwen37-4b-scaling-the-horizon-not-the-parameter
- i-fine-tuned-qwen35-4b-on-3k-browser-trajectories-and-improv
- internlmintern-decision-4b-and-08b
- jev-vs-kev-open-source-jev-alternative-tested-side-by-side
- microsoftfrognano-4b-2609-hugging-face
- model-grafting-turning-qwen35-4b-into-a-causal-encoder-decod
- nuextract3-is-now-available-on-ollama-4b-vlm-for-document-to
- quantize-the-target-quantize-the-drafter-efficient-inference
- tencentevie-preview-45b-hugging-face
- vllm-mlx-vs-ollama-mlx-ran-concurrency-benching-qwen35-4b
Soft Actor-Critic
Mentioned in 10 entries
- adaptive-reinforcement-learning-for-robust-open-quantum-syst
- an-analysis-of-mirror-descent-soft-actor-critic
- dissecting-discrete-soft-actor-critic-limitations-and-princi
- distributional-reinforcement-learning-via-the-cramer-distanc
- quantifying-the-energy-floor-direct-measurement-and-replay-b
- rank-collapse-is-recoverable-growing-q-is-not-out-of-sample-
- reinforcement-learning-position-control-of-a-quadrotor-using
- shared-voxel-map-based-cooperative-indoor-uav-guidance-with-
- toppo-rethinking-ppo-for-multi-task-reinforcement-learning-w
- unthrottling-the-tanh-jacobian-in-sac-a-negative-result-on-b
Claude Fable
Mentioned in 10 entries
- claude-fable-running-comfyui-workflows-through-the-comfy-mcp
- cybersecurity-researchers-complain-that-claude-fables-guardr
- grok-45-delivering-fable-level-performance-at-like-117th-the
- if-claude-fable-stops-helping-you-youll-never-know
- microsoft-restricts-claude-fable-for-employees-over-data-ret
- shall-we-play-a-game---h3---wargames83-ending-spoiler-recrea
- somewhat-humbling-to-have-claude-fable-do-a-final-review-of-
- sqlite-utils-40rc2-mostly-written-by-claude-fable-for-about-
- when-claude-fable-kicks-off-a-workflow-the-tokens-can-go-ver
- wooh-httpsxcomshadcnstatus2064671802509410806s46
Laya
Mentioned in 10 entries
- convaiinnovationslaya-multilingual-non-autoregressive-system
- i-gave-jev-laya-finetuned-modernce-and-qwen35-the-controls-t
- is-typesafe-basedderived-from-work-done-by-the-laya-author
- lavoir-teaching-a-single-pass-decision-encoder-when-and-what
- layacpp-optimized-laya-near-instant-decision-making
- llama-server-add-v1systemone-api-models-laya-julia-1-lev-ope
- ollama-tested-tev1-nimble-and-laya-on-my-pacman-arena-game-s
- stuntd-a-local-jev-compatible-server-on-laya-that-learns-fro
- thanks-for-the-model-we-were-able-to-port-laya-to-coreml-wit
- the-first-model-of-laya-was-created-with-a-single-over-night
GPT-6.1 Sol
Mentioned in 10 entries
- bring-near-astra-intelligence-to-everyday-work-with-gpt-61-s
- gpt-61-sol-is-available-starting-today-to-all-plus-pro-busin
- gpt-61-sol-near-astra-intelligence-for-a-fifth-of-the-price-
- gpt-61-sol-shows-major-improvements-over-gpt-6-sol-in-our-al
- openai-prices-gpt-61-sol-at-21m-input-and-101m-output-tokens
- openais-gpt-61-sol-delivers-astra-like-performance-at-a-dram
- the-bigger-change-is-at-the-low-end-at-low-effort-gpt-61-sol
- ultrafast-is-available-today-for-gpt-6-astra-in-codex-chatgp
- we-comprehensively-benchmarked-gpt-61-sol-on-document-ocr-ta
- were-also-reopening-pro-200-subscriptions-with-continued-acc
Illustrious
Mentioned in 9 entries
- anima-vs-illustrious
- better-indoor-backgrounds-with-illustrious-checkpoints
- does-anyone-have-a-good-example-dataset-for-an-illustrious-c
- i-got-tired-of-writing-massive-prompts-so-i-built-a-local-ra
- illustrious-noobai-style-explorer-5000-danbooru-artist-style
- illustriousnoob-ai-danbooru-tagging-getting-split-up
- is-there-any-way-to-mix-the-style-of-a-sd15-model-with-the-p
- should-i-try-to-convert-a-fp16-illustrious-model-to-fp8
- trying-to-make-an-illustrious-lora-does-anyone-know-of-a-too
Mixture-of-Experts
Mentioned in 9 entries
- cacheable-by-design-training-mixture-of-experts-routers-for-
- enhancing-mixture-of-experts-specialization-via-cluster-awar
- hierarchical-copula-gumbel-top-exorpdfstringkk-routing-two-s
- mixture-of-experts-can-surpass-dense-llms-under-strictly-equ
- mobie-efficient-inference-of-mixture-of-binary-experts-under
- on-the-utility-and-factual-reliability-of-pruned-mixture-of-
- routing-aligned-fine-tuning-for-multilingual-downstream-task
- ubep-re-architecting-expert-parallelism-communication-librar
- wisdom-wireless-sparse-decision-transformer-with-mixture-of-
Gemini 3
Mentioned in 9 entries
- entry-level-guide-to-the-use-of-large-language-models-for-me
- gemini-31-flash-lite-is-now-generally-available-on-gemini-en
- gemini-38-flash-might-be-maxxed-on-other-benchmarks-but-not-
- gemma-4-on-ai-gateway
- google-just-casually-disrupted-the-open-source-ai-narrative
- it-is-going-to-be-like-what-happened-in-coding-as-soon-as-mo
- ockbench-measuring-the-efficiency-of-llm-reasoning
- physical-foundation-models-fixed-hardware-implementations-of
- share-your-gemma-4-builds-or-the-model-variants-youre-traini
Devin
Mentioned in 9 entries
- evinova-a-separate-health-tech-company-within-the-astrazenec
- httpsxcomwaldenyanstatus2052070983083942322s20
- i-have-another-saas-to-kill-will-share-results-of-kill-my-sa
- let-devin-work-the-night-shift-at-140-am-jalagareth-told-dev
- on-frontiercode-11-extended-our-benchmark-for-real-world-eng
- one-of-our-engineers-jhanikhil-got-devin-for-terminal-workin
- qa-with-scott-wu-ceo-of-cognition-which-hit-a-445m-revenue-r
- rivian-and-volkswagen-groups-joint-venture-put-devin-to-work
- try-out-devin-with-the-gpt-55-agent-preview-today-httpsdevin
GANs
Mentioned in 9 entries
- a-large-scale-evaluation-of-text-guided-models-for-facial-ed
- dgs-net-distillation-guided-gradient-surgery-for-clip-fine-t
- maco-gan-manifold-contrastive-adversarial-learning-for-singl
- muflow-leveraging-average-images-for-improving-generalisatio
- ns-net-decoupling-clip-semantic-information-through-null-spa
- seed-to-seed-unpaired-image-translation-in-diffusion-seed-sp
- semi-supervised-bayesian-gans-with-log-signatures-for-uncert
- tabular-gans-for-uneven-distribution
- towards-high-fidelity-face-swapping-a-comprehensive-survey-a
Gemma 4 E4B
Mentioned in 9 entries
- 23-gemma4-e4b-models-compared-with-abliterlitics-the-most-do
- can-you-use-qwen35-4b-gemma-4-e4b-with-z-imageturbo
- chat-uis-with-native-audio-input-for-multimodal-models
- gemma-4e4b-offloads-to-ram-despite-having-just-half-of-vram-
- holy-the-one-and-only-elderplinius-just-dropped-an-unlocked-
- honest-data-ollama-64-toks-vs-llamacppvulkan-16-toks-on-gemm
- i-put-gemma-4-e4b-and-e2b-into-an-e-reader-so-i-can-ask-my-w
- introducing-the-fast-gemma-challenge-with-hugging-face-over-
- mtp-support-for-gemma-4-e2b-and-e4b-assistants-by-max-krasny
FSD
Mentioned in 9 entries
- endlich-konnte-ich-fsd-in-deutschland-testen-wir-sind-fast-e
- first-fsd-in-europe
- former-tesla-data-labelers-say-fsd-relies-on-laborious-mappi
- fsd-is-shockingly-good-i-am-completely-gobsmacked-tesla-team
- fsd-learned-which-spot-at-my-apartment-complex-is-my-assigne
- people-wonder-if-fsd-is-safe-on-narrow-european-roads-well-h
- tesla-self-driving-handles-dutch-roundabouts-well
- today-tesla-fsd-is-9x-safer-than-humans-soon-fsd-will-be-100
- with-fsd-v1432-we-unified-the-model-between-actually-smart-s
UNet
Mentioned in 9 entries
- advanced-flood-prediction-with-physics-guided-deep-learning-
- ciunet-a-hybrid-swin-cnn-unet-for-medical-image-segmentation
- lattice-theory-and-algebraic-models-for-deep-convolutional-l
- optimodnet-a-unet-transformer-hybrid-with-grouped-query-and-
- unet-af-alias-free-unet-architectures
- video-rate-streaming-stylization-on-a-vision-aware-mllm-cond
- vision-transformer-conditioned-unet-for-domain-adaptive-sema
- what-matters-in-virtual-try-off-dual-unet-diffusion-model-fo
- when-geometry-aligns-dihedral-hidden-state-transformations-i
Qwen3-TTS
Mentioned in 9 entries
- adding-emotion-control-tags-to-qwen3-tts
- b10270
- b10311
- free-ai-voice-cloning-with-qwen3-tts-google-colab-notebook-w
- jesus-qwen3-tts-is-fantastic-going-for-full-local-sttttsllm-
- pibot-is-now-running-fully-local-using-parakeet-for-stt-qwen
- qwen3-tts-and-faster-qwen3-tts-on-comfyui
- qwen3-tts-voice-cloning-is-now-in-mainline-llamacpp-the-old-
- ultrafast-qwen3-tts-at-34-ms-time-to-first-audio-handling-10
Diffusion Language Models
Mentioned in 9 entries
- canvasanneal-curriculum-reinforcement-learning-for-diffusion
- dataset-level-metrics-attenuate-non-determinism-a-fine-grain
- expert-choice-routing-enables-adaptive-computation-in-diffus
- learnability-informed-fine-tuning-of-diffusion-language-mode
- memdlm-memory-enhanced-dlm-training
- prefilling-dllm-predictive-prefilling-for-long-context-infer
- prefix-denoising-consistency-test-time-verification-for-diff
- reinforcing-the-generation-order-of-multimodal-masked-diffus
- towards-efficient-reasoning-learning-causal-shortcuts-for-di
Flux Klein
Mentioned in 9 entries
- anyone-using-flux-klein-on-6700xt-or-below-32gb-or-below-ram
- do-you-notice-that-variety-collapses-when-training-style-lor
- flux-klein-is-better-than-any-closed-model-for-image-editing
- how-do-i-replicate-the-dlss-5-output-using-flux-klein
- how-do-you-deal-with-flux-kleins-yellow-tint
- is-there-any-way-to-get-flux-klein-to-not-change-faces-when-
- qwen-edit-vs-flux-klein
- so-how-do-people-fine-tune-the-newer-models-like-flux-klein-
- which-model-would-be-the-best-to-generate-fictional-country-
Composer 2
Mentioned in 9 entries
- composer-25-is-a-significant-step-up-from-composer-2-this-is
- composer-25-is-built-on-the-same-open-source-base-as-compose
- kimi-has-been-the-most-popular-model-on-fireworks-both-out-o
- thanks-for-the-call-out-and-glad-to-see-that-our-blog-inspir
- the-big-lesson-from-training-cursorai-composer-2-models-expl
- the-cursorai-team-shipped-composer-2-and-now-composer-25-on-
- this-weekend-were-doubling-composer-2-usage-in-our-new-inter
- through-the-end-of-this-weekend-we-are-doubling-composer-2-u
- were-launching-kimi-k26-on-fireworks-as-a-day-0-launch-partn
Vision-Language-Action models
Mentioned in 9 entries
- agentic-vla-efficient-online-adaptation-for-vision-language-
- apt-action-expert-pretraining-improves-instruction-generaliz
- deepthinkvla-enhancing-reasoning-capability-of-vision-langua
- drop-then-recovery-how-redundant-are-vision-language-action-
- position-vision-language-action-models-cannot-be-verified-to
- rlinf-vla-a-unified-and-efficient-framework-for-reinforcemen
- sam3d-guided-object-centric-representation-alignment-for-vis
- uaor-uncertainty-aware-observation-reinjection-for-vision-la
- vla-fail-efficient-task-failure-detection-for-finetuned-visi
ColBERT
Mentioned in 9 entries
- clinicalencoder26am-a-multlilingual-diagnosable-colbert-mode
- colbertsar-sparsified-colbert-index-via-product-quantization
- diagnosable-colbert-debugging-late-interaction-retrieval-mod
- eigenli-spectral-approximations-to-late-interaction
- lemur-learned-multi-vector-retrieval
- no-more-k-meanssingle-stage-sparse-coding-for-efficient-mult
- releasing-my-first-kernel-on-huggingface-maxsim-late-interac
- the-amazingly-fast-only-30m-and-extremely-accurate-answerdot
- we-causally-trained-a-lot-of-sota-search-models-internally-s
Parakeet
Mentioned in 9 entries
- b11238
- jesus-qwen3-tts-is-fantastic-going-for-full-local-sttttsllm-
- love-it-ggerganov-s-llamacpp-delivering
- orukeet-multilingual-asr-with-frozen-gabor-kernels
- orukeet-new-asr-model-based-on-parakeet
- pibot-is-now-running-fully-local-using-parakeet-for-stt-qwen
- together-ai-built-the-worlds-fastest-speech-to-text-stack-pa
- voice-agents-get-a-lot-more-interesting-when-they-can-use-th
- what-a-wonderful-project-parakeetcpp-httpsgithubcommudlerpar
Flux 2 Klein
Mentioned in 9 entries
- about-flux-2-klein-img2img-edit-prompting
- creating-character-turnaround-sheets-with-flux-2-klein-in-co
- fizgig-klein-9b-lora-studio-v124---update-targeting-16gb-car
- flux-2-klein-rtx-3060-12gb-fp8-is-almost-same-as-gguf
- flux-2flux-2-klein-transparent-background-lora
- flux2-klein-tiling-upscale-workflow
- hildegard---tiled-upscaling-and-refining-based-on-flux-2-kle
- how-much-is-the-quality-loss-in-flux-2-klein
- how-to-get-klein-to-generate-more-realistic-bodies
CNNs
Mentioned in 9 entries
- ai-generated-image-recognition-via-fusion-of-cnns-and-vision
- bmcr-adaptive-backbone-module-composition-via-reinforcement-
- core-stack-meta-learning-for-deep-stacked-generalization
- does-explainability-transfer-a-controlled-benchmark-of-attri
- lattice-theory-and-algebraic-models-for-deep-convolutional-l
- precise-localization-within-the-gi-tract-by-combining-classi
- season-spectrum-aware-orthogonal-gradient-refinement-for-tra
- texture-representations-in-deep-vision-models-comparing-cnns
- when-is-3d-worth-it-a-resource-performance-frontier-for-cnns
ModernBERT
Mentioned in 9 entries
- b11273
- climate-modernbert-revisiting-corpus-composition-for-domain-
- filter-then-verify-a-multiphase-gnn-and-modernbert-framework
- ibm-granite-just-released-two-multilingual-embedding-models-
- legal-domain-adaptation-of-modern-bert-models
- mogancolbert-tr-a-late-interaction-multi-vector-retrieval-mo
- norberto-a-modernbert-model-trained-for-portuguese-with-331-
- polish-modernbert-the-long-and-short-of-polish-language-unde
- why-advanced-encoders-lag-on-sparse-retrieval-the-answer-and
Segment Anything
Mentioned in 9 entries
- 3d-modeling-and-automated-measurement-of-concrete-cracks-via
- from-pixels-to-concepts-do-segmentation-models-understand-wh
- medsam3-delving-into-segment-anything-with-medical-concepts
- pinpoint-prompting-with-informative-interior-points
- sam-v-geometry-aware-segment-anything-for-multi-view-instanc
- samamba3d-adapting-segment-anything-for-generalizable-3d-seg
- samatcher-co-visibility-modeling-with-segment-anything-for-r
- sarif-segment-anything-for-robust-image-forensics
- temporal-emerged-prompting-for-segment-anything-in-multifram
Laguna XS.2
Mentioned in 9 entries
- congrats-to-poolsideai-on-laguna-xs2-release-day-0-support-i
- first-open-weight-model-from-poolsideai-apache-license-and-a
- laguna-m1xs2-technical-report
- loving-the-latentspacepod-breakdown-of-our-laguna-m1xs2-tech
- poolside-is-hosting-a-2-day-model-research-hackathon-in-lond
- today-poolsideai-is-releasing-laguna-m1-laguna-xs2-our-lates
- today-were-releasing-laguna-xs2-poolsides-first-open-weight-
- today-were-shipping-laguna-m1-and-laguna-xs2-our-first-publi
- us-startup-poolside-debuts-its-first-open-weight-model-lagun
Muse
Mentioned in 9 entries
- i-have-some-questions-for-mark-zuckerberg
- it-seems-like-spacexgrok-and-metamuse-have-started-to-keep-p
- liked-muse-so-i-cut-the-30b-model-in-half-by-width-distilled
- lwiai-podcast-258---opus-55-sol-and-luna-muse-deepseek-v41-f
- meta8217s-muse-ai-sent-a-youtubers-address-to-a-stranger
- sensor-tower-muse-was-downloaded-902k-times-in-the-six-days-
- shopify-plans-to-allow-metas-muse-to-complete-purchases-on-b
- sources-openai-is-developing-features-to-counter-spacexs-gro
- the-dots-are-coming-a-first-experience-in-the-netherlands
GPT 5.6 Sol
Mentioned in 9 entries
- ainews-openai-launches-gpt-56-solterraluna-codex-becomes-cha
- congrats-to-openai-on-the-highly-anticipated-launch-of-gpt-5
- deepseek-v41-flash-is-available-now-on-fireworks-it-is-a-552
- frontier-models-sabotaging-local-ai-implementations
- gpt-56-is-now-available-in-devin-on-fontiercode-11-extended-
- gpt-56-sol-luna-and-terra-now-available-on-ai-gateway
- i-had-a-pretty-similar-experience-a-bunch-of-my-tweets-about
- i-just-ran-qwen-38-27-in-q4-against-gpt-56-sol-high---and-it
- the-cloud-was-always-a-cat-qwen38-max-just-saw-it-first-try-
M3
Mentioned in 9 entries
- chinese-ai-developer-minimax-launches-m3-a-new-coding-model-
- cloud-usage-limits
- introducing-gen-1-slides-an-open-model-that-matches-claude-o
- long-context-sparse-attention-has-a-catch-data-dependent-blo
- minimax-dropped-a-new-attention-architecture-n
- minimaxai-m3-model-is-available-on-ollamas-cloud-in-partners
- see-you-tomorrow-night-come-with-questions
- we-wrapped-a-live-session-on-m3-yesterday-with-the-togetherc
- well-get-into-minimaxai-m3s-model-performance-the-msa-archit
DDPM
Mentioned in 9 entries
- convergence-of-diffusion-models-under-the-manifold-hypothesi
- ddpm-score-matching-and-distribution-learning
- flicker-ddpm-accelerating-denoising-diffusion-via-1f-colored
- improved-denoising-diffusion-probabilistic-models-with-effic
- peel-ddpm-physics-enabled-evidential-learning-for-the-denois
- spatially-adaptive-noise-injection
- supersampling-stable-diffusion-and-more-an-approach-for-inte
- understanding-lora-rank-trade-offs-in-diffusion-model-fine-t
- why-ddim-hallucinates-more-than-ddpm-a-theoretical-analysis-
ResNet
Mentioned in 9 entries
- automated-brain-tumor-detection-in-mri-images-using-cnn-and-
- automated-ecg-interval-measurement-and-wave-delineation-usin
- benchmarking-resnet-backbones-in-rt-detr-impact-of-depth-and
- benchmarking-resnet-for-short-term-hypoglycemia-classificati
- difta-3d-depth-consistent-instance-level-feature-transfer-an
- effect-of-demographic-bias-on-skin-lesion-classification
- phase-transition-frequency-as-a-training-time-predictor-of-t
- spatiality-frequency-domain-video-forgery-detection-system-b
- student-capacity-moderates-knowledge-distillation-effectiven
Sakana Fugu
Mentioned in 9 entries
- how-does-it-work-sakana-fugu-is-itself-an-llm-trained-to-cal
- one-of-my-favorite-things-about-sakana-fugu-is-the-recursive
- sakana-ai-ceo-david-ha-hardmaru-appeared-on-tbs-cross-digs-1
- sakana-ai-sakana-ai-sakana-chat-sakana-marlin-sakana-fugu
- sakana-fugu-httpssakanaaifugu-releasejapanese-sakana-aisakan
- sakana-fugu-technical-report
- were-hiring-software-engineer-rd-infrastructure-and-platform
- were-launching-the-beta-for-our-new-commercial-ai-product-sa
- weve-been-using-sakana-fugu-internally-for-our-own-research-
GPT-Rosalind
Mentioned in 9 entries
- gpt-rosalind-our-life-sciences-model-series-is-optimized-for
- introducing-gpt-rosalind-for-life-sciences-research
- introducing-gpt-rosalind-our-frontier-reasoning-model-built-
- introducing-new-capabilities-to-gpt-rosalind
- openai-launches-gpt-rosalind-an-ai-model-for-life-sciences-r
- openai-says-it-has-briefed-the-white-house-on-its-new-biodef
- openai-starts-offering-a-biology-tuned-llm
- strengthening-societal-resilience-with-rosalind-biodefense
- were-bringing-new-capabilities-to-gpt-rosalind-a-model-serie
Qwen3.8-2.4T-A95B
Mentioned in 9 entries
- deploying-qwen38-24t-a95b-on-amazon-sagemaker-hyperpod-with-
- experiment-qwen38-24t-a95b-running-locally-on-an-rtx-5090-rt
- from-idea-to-implementation-in-one-go-max-level-intelligence
- max-level-intelligence-one-api-call-away-qwen38-24t-a95b-is-
- qwen38-24t-a95b-is-live-on-together-ai-the-qwen-teams-new-op
- qwen38-24t-a95b-is-now-live-on-fireworks-with-day-0-support-
- qwen38-24t-a95b-is-now-live-on-together-ai-the-qwen-teams-la
- serve-qwen38-24t-a95b-a-24t-parameter-model-with-configurabl
- together-serverless-inference-gives-developers-a-managed-hig
ChatGPT Pro
Mentioned in 9 entries
- an-obvious-way-to-release-mythos-class-models-with-uncertain
- bet-this-happens-with-navier-stokes-and-its-going-to-be-some
- did-i-just-pay-100-for-15-days-of-chatgpt-pro
- did-the-100-plan-affect-the-gpt-54-pro-model
- me-waiting-for-gpt-6-astra-access-on-my-chatgpt-pro-subscrip
- openai-debuts-personal-finance-tools-for-us-chatgpt-pro-user
- openai-says-it-will-pause-new-200month-chatgpt-pro-subscript
- show-us-the-codex-pets-you-hatched-use-hatch-to-create-your-
- starting-today-you-can-now-use-your-chatgpt-subscription-dir
Copilot
Mentioned in 9 entries
- accenture-copilot-employee-rollout-results
- agi-ftw-ft
- artificial-intelligence-in-sports-insights-from-a-quantitati
- friends-outside-of-tech-lol-copilot-is-dumb---friends-in-tec
- gemini-is-in-danger-of-going-full-copilot
- hit-my-claude-and-copilot-limits-three-times-in-each-of-the-
- microsoft-releases-web-iq-a-search-service-for-ai-agents-tha
- novee-debuts-agentic-fix-pushing-pentest-findings-into-claud
- tcm-serve-modality-aware-scheduling-for-multimodal-large-lan
Command A+
Mentioned in 9 entries
- cohere-is-on-such-a-great-open-source-trajectory-lately-beau
- command-a-from-cohere-is-out-now-its-our-best-model-yet-and-
- command-a-is-available-on-huggingface-with-w4a4-quantization
- command-a-sets-a-new-high-for-coheres-machine-translation-ca
- introducing-cohere-command-a-weve-created-our-most-powerful-
- open-source-nvidia
- the-improvements-run-wide-across-all-major-european-language
- the-story-gets-bigger-beyond-europe-command-a-makes-major-ga
- wait-did-cohere-just-release-command-a-models-under-apache-2
ChatGPT Images 2.5
Mentioned in 9 entries
- chatgpt-images-25-in-the-wild-a-launch-period-dataset-and-de
- chatgpt-images-25faster-sharper-smarter-with-better-tools-fo
- chatgpt-sketch-turns-your-bad-drawings-into-detailed-ai-imag
- from-advertised-improvements-to-measured-capabilities-evalua
- heres-a-blender-model-of-a-pluribus-themed-faberg-egg-i-had-
- images-25-is-here-i-dont-think-it-can-solve-super-difficult-
- introducing-chatgpt-images-25
- openai-launches-chatgpt-images-25-which-it-says-cuts-image-g
- openais-chatgpt-images-gets-faster-with-sharper-details-and-
Gemini 3 Flash
Mentioned in 9 entries
- ai-researchers-ran-15-day-simulations-of-worlds-governed-by-
- gemini-3-flash-previewcloud-403-forbidden-error
- googles-new-gemini-35-flash-is-the-clear-leader-on-the-intel
- my-notes-on-gemini-35-flash---3x-the-price-of-gemini-3-flash
- ocrllamaindex
- ollama-cloud---which-llm-is-faster-than-gemini-3-flash-now
- study-shows-ai-chooses-nuclear-war-in-crisis
- the-best-raw-frontier-model-for-document-parsing-is-gemini-3
- this-paper-by-researchers-from-mit-and-stanford-finds-that-m
Claude Sonnet
Mentioned in 9 entries
- build-ai-agents-for-business-intelligence-with-amazon-bedroc
- claudes-voice-mode-is-now-available-for-opus-and-sonnet
- copilot-just-9xd-sonnet-and-27xd-opus-and-teams-have-no-idea
- if-llama-4-didnt-fail-if-microsoft-had-pulled-sydney-after-t
- long-review-qwen-38-27b-is-very-good-at-tapping-into-its-rea
- my-experience-with-qwen38
- the-building-ai-agents-with-mongodb-and-langgraph-skill-badg
- trying-to-build-a-local-claude-sonnet-like-cli-coding-setup-
- which-local-models-handle-structured-extraction-well-i-moved
Claude Max
Mentioned in 9 entries
- anthropic-launches-claude-opus-5-which-it-says-comes-close-t
- cant-merge-claude-max-subscription-opus-48-with-zai-glm-521m
- fable-5-is-extended-through-july-12th-the-number-of-people-t
- fable-gets-another-bump
- hired-a-dev-instead-of-buying-claude-subscription-now-he-wan
- i-actually-cancelled-my-claude-max-subscription-well-downgra
- i-switched-from-anthropics-200mo-claude-max-to-ollama-clouds
- up-until-yesterday-our-entire-mts-team-has-operated-under-th
- weve-got-13-days-to-burn-as-much-tokens-as-humanly-possible-
Gemini 3.8 Live
Mentioned in 9 entries
- check-out-this-weeks-updates-and-releases-gemini-38-flash-tt
- gemini-38-live-is-rolling-out-to-consumers-in-search-live-de
- gemini-38-live-with-live-avatar-gives-google8217s-ai-a-face
- gemini-live-audio
- google-updates-gemini-38-live-with-live-avatar-an-animated-a
- googles-new-speech-model-gemini-38-live-supports-real-time-r
- introducing-our-most-advanced-gemini-audio-models-yet-gemini
- its-time-for-our-end-of-week-recap-gemini-38-live-and-38-liv
- power-your-agents-gemini-38-live-with-live-avatar-is-now-gen
VLMs
Mentioned in 8 entries
- can-vlms-unlock-semantic-anomaly-detection-a-framework-for-s
- dso-direct-steering-optimization-for-bias-mitigation
- multimodal-unlearning-across-vision-language-video-and-audio
- rethinking-visual-privacy-a-compositional-privacy-risk-frame
- safe-task-planning-with-long-term-graph-memory-for-embodied-
- sallie-safeguarding-against-latent-language-image-exploits
- visual-compliance-via-executable-safety-rule-entailment
- vlms-may-not-globally-enhance-human-alignment-over-llms-duri
OpenClaw
Mentioned in 8 entries
- ask-now-use-later-benchmarking-the-proactivity-gap-in-long-l
- deepseek-v4-flash-is-now-available-on-ollamas-cloud-hosted-i
- liteparse-should-be-the-default-document-parser-you-use-with
- migrate-from-openclaw-or-hermes
- openforgerl-train-harness-native-agents-in-any-environment
- skillclaw-let-skills-evolve-collectively-with-agentic-evolve
- skillrae-agent-skill-based-context-compilation-for-retrieval
- tried-running-ollama-with-openclaw-but-ran-out-of-ram-lookin
Adam
Mentioned in 8 entries
- a-trust-region-framework-for-moment-estimation
- adam-hnag-a-convergent-reformulation-of-adam-with-accelerate
- anon-extrapolating-optimizer-adaptivity-across-the-real-spec
- efficient-long-horizon-learning-for-learned-optimization
- flowadam-implicit-regularization-via-geometry-aware-soft-mom
- phases-of-muon-when-muon-eclipses-signsgd
- powerstep-memory-efficient-adaptive-optimization-via-ellp-no
- tensorion-a-tensor-aware-generalization-of-the-muon-optimize
Swin Transformer
Mentioned in 8 entries
- gated-swinrmt-unifying-swin-windowed-attention-with-retentiv
- geoformer-a-lightweight-swin-transformer-for-joint-building-
- multimodal-skin-lesion-classification-with-swin-transformer-
- on-privacy-preserving-image-transmission-in-low-altitude-net
- parallel-swin-transformer-enhanced-3d-mri-to-ct-synthesis-fo
- spatial-frequency-gated-swin-transformer-for-remote-sensing-
- swinifs-landmark-guided-swin-transformer-for-identity-preser
- swintextunet-integrating-clip-based-text-guidance-into-swin-
Random Forest
Mentioned in 8 entries
- bring-your-own-model-with-amazon-sagemaker-ai-script-mode-in
- feature-space-selection-and-heterogeneous-effect-estimation-
- how-many-trees-in-a-random-forest-a-revisited-approach-with-
- integrating-saint-with-tree-based-models-a-case-study-in-emp
- introducing-tabfm-in-bigquery-predictive-analytics-reimagine
- lightweight-lidar-based-cone-detection-framework-using-rando
- towards-the-automated-segmentation-of-epicardial-and-mediast
- when-clean-signals-are-not-enough-detecting-structural-ambig
Gaussian Splatting
Mentioned in 8 entries
- a-comparative-evaluation-of-geometric-accuracy-in-nerf-and-g
- fixanything-3d-consistent-rendering-refinement-via-video-gen
- medgs-gaussian-splatting-for-multi-modal-3d-medical-imaging
- pagas-pixel-aligned-1dof-gaussian-splatting-for-depth-refine
- resplat-learning-recurrent-gaussian-splatting
- rpc-gs-gaussian-splatting-with-native-rpc-rendering-for-sate
- splatography-sparse-multi-view-dynamic-gaussian-splatting-fo
- vs-splat-voxel-selective-feed-forward-gaussian-splatting-for
ERNIE-Image
Mentioned in 8 entries
- ernie-image-comics-w-sample-prompt-great-ability-to-track-mu
- ernie-image-is-now-in-comfyui-an-open-source-8b-dit-text-to-
- ernie-image-technical-report
- ernie-image-turbo-is-capable-of
- get-started-in-comfyui-1-update-comfyui-to-0191-desktop-and-
- ostris-ai-toolkit-has-day-zero-support-for-training-loras-on
- unlike-zit-ernie-image-seems-to-be-really-good-for-lora-trai
- we-may-have-a-new-sota-open-source-model-ernie-image-compari
RWKV
Mentioned in 8 entries
- belief-state-rwkv-for-reinforcement-learning-under-partial-o
- compass-degradation-simulated-reciprocal-learning-with-light
- efficient-rwkv-based-representation-learning-for-3d-point-cl
- kan-we-flow-advancing-robotic-manipulation-with-3d-flow-matc
- mg-rwkv-multi-grained-context-aware-rwkv-for-temporal-forger
- multigrain-aware-semantic-prototype-scanning-and-tri-token-p
- scrwkv-ultra-compact-structure-calibrated-vision-rwkv-for-to
- triplet-block-diffusion-rwkv
Vision Mamba
Mentioned in 8 entries
- beyond-zoh-advanced-discretization-strategies-for-vision-mam
- dsvm-unet-enhancing-vm-unet-with-dual-self-distillation-for-
- event-adaptive-state-transition-and-gated-fusion-for-rgb-eve
- fractalmamba-scaling-vision-mamba-across-resolutions-via-hil
- mambampd-a-mamba-driven-segmentation-framework-for-marine-po
- mambapanoptic-a-vision-mamba-based-structured-state-space-fr
- norm-or-direction-decoding-vision-mambas-for-high-resolution
- vim-q-scalable-algorithm-hardware-co-design-for-vision-mamba
o1-preview
Mentioned in 8 entries
- context-from-the-time-of-the-o1-preview-launch-httpswwwoneus
- evaluation-of-openai-o1-opportunities-and-challenges-of-agi
- i-demonstrated-the-incredible-power-of-o1-preview-reasoning-
- id-been-part-of-openai-early-tester-group-for-gpt-55-i-belie
- it-is-less-than-a-decade-since-the-development-of-the-transf
- its-funny-how-much-the-whole-strawberry-thing-which-turned-o
- strawberry-was-the-popular-code-word-for-reasoning-as-implem
- the-great-thing-is-that-the-names-are-so-baffling-that-the-m
Midjourney
Mentioned in 8 entries
- disponibilizei-meu-workflow-chroma-v48-dc-v48-best-midjourne
- do-we-have-any-good-loras-fine-tune-for-style-similar-to-nij
- findings-of-the-counter-turing-test-ai-generated-image-detec
- image-generation-models-running-locally-on-limited-resources
- mj-style-distilled-206
- this-is-an-interesting-point-in-the-new-ted-chiang-piece-no-
- what-image-model-should-i-use-as-somebody-who-likes-the-aest
- what-underrated-ai-tools-have-actually-made-you-more-product
Neural Radiance Fields
Mentioned in 8 entries
- asyncevgs-asynchronous-event-assisted-gaussian-splatting-for
- beyond-a-single-light-a-large-scale-aerial-dataset-for-urban
- fast-and-lightweight-novel-view-synthesis-with-differentiabl
- fixanything-3d-consistent-rendering-refinement-via-video-gen
- gor-is-3d-gaussian-object-removal-in-the-intrinsic-space
- pandora-casual-hdr-radiance-acquisition-of-indoor-scenes-for
- render-fm-feedforward-model-for-real-time-photorealistic-vol
- routebridge-reliability-routed-bidirectional-distillation-be
GPT-Realtime-2
Mentioned in 8 entries
- ainews-gpt-realtime-2--translate-and--whisper-new-sota-realt
- introducing-gpt-realtime-2-in-the-api-our-most-intelligent-v
- our-new-voice-models-are-now-available-in-the-realtime-api-g
- people-are-really-starting-to-use-voice-to-interact-with-ai-
- process-speech-as-speech-with-gpt-realtime-2-or-gemini-31-li
- saw-this-and-thought-yes-chatgpt-voice-mode-is-going-to-stop
- try-grok-voice
- whos-adding-this-to-reachy-mini
Ideogram
Mentioned in 8 entries
- a-helpful-little-tip-to-help-deal-with-the-ideogram-model-ce
- alright-ill-be-that-guy-what-are-your-ideogram-settings-the-
- i-didnt-expect-ideogram-to-be-so-good
- ideogram-making-2-horrible-precedent-and-we-need-to-oppose-t
- ideogram-model---lora
- lodestone-is-thinking-about-training-ideogram-prove-him-its-
- sorry-not-sorry-ideogram-jailbroken-in-1-easy-step
- trained-my-firsts-loras-on-ideogram-and-it-is-seriously-a-ga
Mamba-2
Mentioned in 8 entries
- and-another-open-weight-release-nemotron-3-ultra-has-an-ultr
- b10164
- b10632
- compiler-first-state-space-duality-and-portable-o1-autoregre
- density-field-state-space-models-1-bit-distillation-efficien
- detection-vs-execution-single-bucket-probes-miss-half-the-ma
- heed-density-weighted-residual-alignment-for-hybrid-vision-l
- mambanetburst-direct-byte-level-network-traffic-classificati
DeepSeek R1
Mentioned in 8 entries
- deepseek-improved-the-code-and-said-nothing-happened-in-tian
- deepseek-r1-community-edition-is-dogshit
- deepseek-v4-the-most-expected-open-source-model-ever-release
- deepseeks-new-ai-is-a-game-changer
- did-they-shut-down-deep-seek-cloud-for-free-users
- openai-says-its-jalapeo-chip-delivered-15x-19x-more-ai-work-
- really-cool-that-mimo-v26-has-an-official-distill-qwen-9b-va
- vera-rubin-nvl72-vs-gb200-nvl72-inference-tco-architecture-a
HiDream-O1-Image
Mentioned in 8 entries
- artificial-analysis-needs-to-address-hidream-01-benchmarks
- comfyui-hidream-text-image-and-image-edit-templates---multip
- comfyui-support-for-hidream-01-image-released
- hidream-o1-image---a-pixel-space-model-no-need-for-vae-8b-pa
- hidream-o1-image-a-natively-unified-image-generative-foundat
- hidream-o1-image-dev-the-showcase-doesnt-match-reality
- hidream-o1-image-internal-prompt
- hidream-studio-v01-has-been-released-it-is-fast-and-powerful
Vision-Language-Action (VLA)
Mentioned in 8 entries
- dial-decoupling-intent-and-action-via-latent-world-modeling-
- guidedvla-specifying-task-relevant-factors-via-plug-and-play
- latent-reasoning-vla-latent-thinking-and-prediction-for-visi
- let-it-be-simple-one-step-action-generation-for-vision-langu
- priorvla-prior-preserving-adaptation-for-vision-language-act
- rise-self-improving-robot-policy-with-compositional-world-mo
- rl2-vla-adaptive-rl-latent-compositional-steering-with-test-
- safealign-vla-a-negative-enhanced-safe-alignment-framework-f
V-JEPA
Mentioned in 8 entries
- a-slightly-improved-dvd-jepa-demo-p
- applying-jepa-style-predictive-learning-to-ja4-derived-netwo
- hpc-enabled-video-based-coastal-wave-parameter-estimation-us
- sim-to-real-traffic-scene-understanding-by-decoupling-semant
- unifying-semantic-priors-and-high-frequency-traces-enhancing
- v-jepa-policy-building-effective-world-action-models-on-pred
- vista-technical-report-for-the-ego4d-short-term-object-inter
- wa-jepa-rethinking-the-video-jepa-paradigm-for-world-action-
AlphaEarth
Mentioned in 8 entries
- aef-econ-toward-plug-and-play-socioeconomic-foundation-embed
- alphaearth-distinguishes-cities-but-compresses-urban-variati
- beacon-behavioral-and-semantic-enrichment-of-alphaearth-embe
- characterizing-alphaearth-embedding-geometry-for-agentic-env
- evaluating-alphaearth-foundations-embeddings-for-wildfire-su
- from-landslide-conditioning-factors-to-satellite-embeddings-
- slum-detection-and-density-mapping-with-alphaearth-foundatio
- utilizing-earth-foundation-models-to-enhance-the-simulation-
Large Vision-Language Models
Mentioned in 8 entries
- beyond-attention-imbalance-mitigating-hallucinations-via-spe
- echotrail-gui-building-actionable-memory-for-gui-agents-via-
- gaia-a-data-flywheel-system-for-training-gui-test-time-scali
- gaussian-mixture-modeling-for-event-aware-visual-allocation-
- hyperbolic-and-evidence-prioritized-experts-for-large-vision
- spatialforge-bootstrapping-3d-aware-spatial-reasoning-from-o
- to-see-is-not-to-learn-protecting-multimodal-data-from-unaut
- visual-sensitivity-is-not-claim-retractability-persistence-a
ESMFold2
Mentioned in 8 entries
- alexrives-on-ai-for-science-with-latentspacepod-breaking-dow
- announcing-esmfold2-our-new-state-of-the-art-structure-predi
- biohubs-protein-world-model-esmc-6b-esmfold2-68b-proteins-11
- esmfold2-is-a-state-of-the-art-folding-model-its-crazy-good-
- esmfold2-the-bitter-lesson-is-coming-for-proteins---alex-riv
- today-was-a-massive-day-for-protein-engineering-esmfold2-dro
- today-were-announcing-esmfold2-an-open-scientific-engine-to-
- took-me-a-while-to-figure-out-what-all-the-esmfold2-rage-was
Step 3.7 Flash
Mentioned in 8 entries
- many-research-labs-only-consider-inference-efficiency-after-
- more-time-to-build-with-step-37-flash-in-partnership-with-st
- other-active-promotions---free-models-solar-pro-4-1-week-hy3
- our-users-love-stepfunai-models-and-this-new-release-packs-a
- run-step-37-flash-on-nvidia-gpus-with-enterprise-ready-multi
- step-37-flash-free-for-30-days-for-hermes-agent-users-what-c
- step-37-flash-is-now-free-for-30-days-via-nous-portal-it-is-
- weve-gotten-some-great-medium-sized-models-lately-dsv4-flash
Inkling-Small
Mentioned in 8 entries
- inkling-small-2-weeks-after-inkling-nearly-as-good-as-inklin
- inkling-small-by-thinkingmachines
- inkling-small-is-now-live-on-together-ai-thinkymachines-new-
- now-suddenly-too-many-choices-for-dgx-spark-with-qwen-35-122
- quick-accurate-summaries-make-httpshuggingfacecopapers-an-ev
- seta-got-accepted-to-neurips-2026-we-released-the-full-train
- thinking-machines-just-released-inkling-small-276b-total-12b
- together-ai-gives-developers-a-high-throughput-production-pa
GRPO
Mentioned in 8 entries
- advancing-entropy-level-credit-assignment-in-rlvr-via-proxim
- alleviating-sparse-rewards-by-modeling-step-wise-and-long-te
- beyond-mode-collapse-distribution-matching-for-diverse-reaso
- devis-grpo-unleashing-grpo-on-dynamic-extreme-view-synthesis
- grape-let-grpo-supervise-query-rewriting-by-ranking-for-retr
- grpo-tta-test-time-visual-tuning-for-vision-language-models-
- how-much-online-rl-is-enough-informative-rollouts-for-offlin
- when-synthetic-speech-is-all-you-have-better-call-grpo
ChatGPT Plus
Mentioned in 8 entries
- 2-separate-gpt-plus-sub-or-business-of-2-for-2-users
- claude-pro-feels-much-more-dynamic-and-visual-than-chatgpt-p
- evaluation-of-medical-vision-language-models-hulumed-and-med
- openai-and-malta-partner-to-bring-chatgpt-plus-to-all-citize
- openai-partners-with-maltas-ai-for-all-initiative-to-give-ci
- openai-restores-5-hour-codex-and-work-limits-for-chatgpt-plu
- ran-chatgpt-plus-and-claude-pro-side-by-side-for-30-days-her
- starting-today-you-can-now-use-your-chatgpt-subscription-dir
AlphaZero
Mentioned in 8 entries
- analysis-of-alphazero-training-data-d
- belief-guided-decision-making-with-uncertainty-gating-in-the
- engineering-efficient-self-play-chess-search-replay-and-thro
- in-2016-david-wrote-in-his-private-journal-that-the-reason-t
- learning-to-run-power-networks-effective-alphazero-inspired-
- maple-multi-state-aggregated-policy-evaluation-for-alphazero
- what-to-expect-from-alphazeros-value-predictions-d
- when-search-teaches-style-causal-internalization-of-tactical
GPT-Image-2
Mentioned in 8 entries
- ainews-openai-launches-gpt-image-2
- editable-visual-design
- given-how-good-gpt-image-2-is-its-an-absolute-puzzle-openai-
- gpt-image-2-does-better-with-small-details-and-text-but-its-
- gpt-image-2-in-the-wild-a-twitter-dataset-of-self-reported-a
- textrich-a-multi-domain-benchmark-for-detecting-ai-generated
- the-president-test-with-gpt-image-2
- when-the-forger-is-the-judge-gpt-image-2-cannot-recognize-it
Sonnet 5
Mentioned in 8 entries
- anthropic-releases-sonnet-55-saying-it-generates-outputs-30-
- deepseek-v4-flash-ga-ranks-the-same-as-sonnet-5-and-grok-45-
- on-frontiercode-extended-our-benchmark-for-real-world-engine
- open-source-tax-engine-outperforming-fable-5-and-gpt-sol
- sonnet-5-is-here-this-is-going-to-support-better-long-runnin
- vercel-cloudflare-and-others-quickly-add-jev-as-it-makes-ai-
- we-analyzed-glm-52-vs-sonnet-5-for-software-engineering-task
- while-the-token-prices-for-sonnet-55-remain-the-same-as-sonn
Deepseek V4
Mentioned in 8 entries
- b10655
- deepseek-v4-flash-105-ts-on-two-nvidia-4090d-48g-ada-in-vllm
- deepseek-v4-works-more-thoroughly-than-other-open-source-mod
- fireworksaihq-deeply-believes-in-delivering-the-frontier-qua
- mention-if-a-new-model-is-a-finetune
- moe-cpu-offload-benchmark-on-deepseek-v4gemma4qwengpt-oss-te
- my-first-week-on-ollama
- with-release-of-deepseek-v4-i-wanted-see-how-the-model-sizes
Gemini 3.7 Flash
Mentioned in 8 entries
- gemini-37-flash-is-now-available-in-devin-on-frontiercode-11
- google-ai-pro-and-ultra-subscribers-can-experience-37-flash-
- google-launches-gemini-37-flash-for-coding-ai-agent-projects
- llm-gemini-033
- lwiai-podcast-255---gemini-37-jalapeo-qwen-38-drones
- new-langchain-oss-release
- our-most-intelligent-workhorse-model-yet-for-coding-and-agen
- within-devin-gemini-37-flash-performs-particularly-well-on-t
Gemini 3.1 Flash TTS
Mentioned in 8 entries
- gemini-31-flash-tts
- gemini-31-flash-tts-the-next-generation-of-expressive-ai-spe
- google-rolls-out-gemini-31-flash-tts-a-text-to-speech-model-
- googles-gemini-31-flash-tts-model-offers-unparalleled-contro
- guide-to-prompting-gemini-31-flash-tts-text-to-speech
- the-example-prompt-for-googles-new-gemini-flash-tts-text-to-
- today-we-launched-gemini-31-flash-tts-our-most-expressive-an
- what-a-week-heres-everything-we-shipped-gemini-31-flash-tts-
GPT-5.5-Cyber
Mentioned in 8 entries
- openai-expands-daybreak-with-patch-the-planet-and-full-gpt-5
- openai-introduces-gpt55cyber-for-high-impact-cybersecurity-r
- openai-unveils-an-updated-gpt-55-cyber-model-launches-the-pa
- openais-new-security-model-is-for-critical-cyber-defenders-o
- scaling-trusted-access-for-cyber-with-gpt-55-and-gpt-55-cybe
- we-want-to-help-all-companies-be-secure-working-with-the-usg
- wed-like-to-help-companies-secure-themselves-and-we-think-it
- were-starting-rollout-of-gpt-55-cyber-a-frontier-cybersecuri
DeepSeek v4 Flash
Mentioned in 8 entries
- 2-bit-deepseek-v4-flash-inference-experts-w13-iq2xxs-w2-q2k-
- btw-it-looks-like-running-the-new-deepseek-v4-flash-through-
- cactus-needle-3-a-sliceable-8-29mb-automation-foundation-mod
- deepseek-v4-flash-for-ds4-dwarfstar-gguf-w-dspark-mtp-head
- deepseek-v4-flash-vs-qwen36-27b-35-122b-and-gemma-4-31b-benc
- deepseek-v4-flash-with-local-inference-after-24h-of-playing-
- local-agentic-coding-benchmark-qwen-38-27b-in-many-weights-q
- welcome-to-ds4-a-specialized-inference-engine-for-deepseek-v
RNN
Mentioned in 7 entries
- comparative-analysis-of-liquid-neural-networks-and-lstm-for-
- decoding-high-dimensional-finger-motion-from-emg-using-riema
- frame-forecasting-in-cine-mri-using-the-pca-respiratory-moti
- geometry-induced-long-range-correlations-in-recurrent-neural
- i-know-its-self-serving-to-say-but-man-i-wouldve-killed-for-
- rgnn-model-for-fraud-detection-isnt-performing-wellr
- scalable-multi-gpu-simulation-of-3d-multicellular-growth-wit
GPT-3.5
Mentioned in 7 entries
- from-gpt-3-to-gpt-5-mapping-their-capabilities-scope-limitat
- i-just-remembered-this-thread-where-i-was-playing-with-gpt-3
- it-is-less-than-a-decade-since-the-development-of-the-transf
- meh-on-the-lem-test-as-well-httpsxcomemollickstatus201947831
- seeing-like-an-ai-how-llms-apply-and-misapply-wikipedia-neut
- the-great-thing-is-that-the-names-are-so-baffling-that-the-m
- the-second-most-important-release-of-the-llm-era-after-gpt-3
AlphaGo
Mentioned in 7 entries
- after-alphago-the-skill-of-human-go-players-noticeably-impro
- announcing-our-partnership-with-the-republic-of-korea
- eladgil-bs-attention-was-born-in-montral-pytorch-in-nyc-alph
- people-compared-gpt-54s-solution-of-erdos-1196-to-alphagos-m
- seems-plausible-people-will-eventually-periodize-ai-history-
- ten-years-ago-alphagos-move-37-shocked-the-world-it-wasnt-in
- the-current-fear-is-is-that-ai-homogenizes-culture-and-turns
SWE-1.6
Mentioned in 7 entries
- or-try-swe-16-in-windsurf-today-httpswindsurfcom
- silicon-valley-is-quietly-running-on-chinese-open-source-ai-
- swe-16-is-free-for-everyone-in-windsurf-for-the-next-3-month
- swe-16-is-lightning-fast-heres-what-950-toks-feels-like---av
- were-releasing-swe-16-our-best-model-in-both-intelligence-mo
- weve-teamed-up-with-cerebras-to-offer-free-windsurf-plans-fo
- with-swe-16-weve-made-significant-progress-on-intelligence-p
Grok 4.20
Mentioned in 7 entries
- both-gemini-and-grok-are-underrated-i-used-to-champion-gemin
- grok-420-hitting-83-on-non-hallucination-values-truth-claude
- grok-420-just-claimed-1-on-ifbench-artificial-analysis---the
- grok-is-the-ai-to-use-if-you-value-truth
- try-grok-420-on-ai-gateway
- try-it
- xai-models-now-support-zero-data-retention-zdr-on-openrouter
LightGBM
Mentioned in 7 entries
- automated-detection-of-dosing-errors-in-clinical-trial-narra
- benchmarking-lightgbm-and-bilstm-for-sentiment-analysis-on-i
- benchmarking-logistic-regression-svm-and-lightgbm-against-bi
- ema-fs-accelerating-gbdt-training-via-gain-informed-feature-
- ensemble-deep-learning-models-for-early-detection-of-meningi
- integrating-saint-with-tree-based-models-a-case-study-in-emp
- omnitabbench-mapping-the-empirical-frontiers-of-gbdts-neural
Llama 4
Mentioned in 7 entries
- entry-level-guide-to-the-use-of-large-language-models-for-me
- how-are-you-running-llama-33-4-in-ollama-from-24gb-cards-up-
- i-think-muse-spark-came-in-far-better-than-most-were-expecti
- if-llama-4-didnt-fail-if-microsoft-had-pulled-sydney-after-t
- in-the-grand-narrative-of-meta-x-ai-we-saw-the-flop-llama-4-
- meta-returns-to-open-weights-muse-glimmer-its-first-open-wei
- metas-new-model-is-muse-spark-and-metaai-chat-has-some-inter
FLUX.2 Klein 9B
Mentioned in 7 entries
- flux-2-klein-9b-produces-absolutely-awful-and-ugly-skin-text
- flux-identity-adjustor-node-for-flux2-klein-9b-model
- flux2-klein-9b-schematic-lora---depth-normal-pose-and-segmen
- flux2-klein-identity-feature-transfer-advanced
- slay-the-spire-2---flux2-klein-9b-style-loras
- transformed-my-office-vibe-with-flux2-klein-9b-with-lora-bef
- wireframe---flux2-klein-9b-style-lora
ZIT
Mentioned in 7 entries
- ernie-shows-some-strength-in-infographic-but-yes-in-photorea
- kelin9bt-vs-ernieit-vs-zit-fft-analysis-of-artifacts
- recreating-80s-and-90s-anime-style-with-zit-and-ltx-23
- testing-zit-and-flux-1-with-nvidia-pid-pixel-diffusion-decod
- zib-zit-hand-over
- zit-better-then-qwen
- zit-vs-zib-settings-question
Ernie
Mentioned in 7 entries
- 2-the-same-message-applies-to-several-models-chroma-z-image-
- 3-the-same-message-applies-to-several-models-chroma-z-image-
- 4-a-mesma-mensagem-aplica-se-a-vrios-modelos-chroma-z-image-
- 5-the-same-message-applies-to-several-models-chroma-z-image-
- any-controlnet-for-ernie
- ernie-image-character-loras-any-luck
- ernie-shows-some-strength-in-infographic-but-yes-in-photorea
Qwen3-4B
Mentioned in 7 entries
- 400-llm-agents-living-in-a-2004-era-mmo-server-all-local-on-
- can-a-4b-local-model-actually-feel-like-an-ai-assistant
- can-a-cacheable-decision-model-follow-rules
- post-training-ternarization-of-qwen3-4b-capability-effective
- qu-nlp-at-archehr-qa-2026-two-stage-qlora-fine-tuning-of-qwe
- we-distilled-23m-claude-fable-5-reasoning-traces-into-qwen3-
- which-small-model-works-best-with-ollama-on-a-raspberry-pi-5
MAE
Mentioned in 7 entries
- low-accuracy-50-with-ssl-byolmaevicreg-on-hyperspectral-crop
- mae-based-self-supervised-pretraining-for-data-efficient-med
- medfedpure-a-medical-federated-framework-with-mae-based-dete
- plankformer-robust-plankton-instance-segmentation-via-mae-pr
- rethink-mae-with-linear-time-invariant-dynamics
- self-distillation-of-hidden-layers-for-self-supervised-repre
- steerable-visual-representations
Gemma 3
Mentioned in 7 entries
- are-1b-llms-going-away-in-2026
- evaluation-of-medical-vision-language-models-hulumed-and-med
- googles-deepmind-just-released-new-4b-and-27b-medgemma-model
- i-built-a-tpu-native-medical-qa-fine-tuning-pipeline-using-g
- open-source-local-app-that-uses-ollama-gemma-3-to-analyze-wh
- recurrence-is-not-enough-causally-validating-multilingual-sa
- we-benchmarked-translategemma-against-5-other-llms-on-subtit
Stable Diffusion 1.5
Mentioned in 7 entries
- benchmarking-local-stable-diffusion-15-generations-on-iphone
- best-way-to-generate-ai-images-locally-on-amd-rx-9070-xt
- looking-for-a-workflow-for-img2img-to-upscale-detail-and-sha
- running-modern-ai-image-models-on-a-gtx-1060-6gb-a-practical
- sd-15-forgeui-advice-needed
- struggling-to-get-clean-pixel-art-from-sd15-lora-what-am-i-m
- these-artifacts-of-the-new-chatgpt-images-20-remind-me-of-th
Video Diffusion Transformers
Mentioned in 7 entries
- attentionbender-manipulating-cross-attention-in-video-diffus
- fis-dit-breaking-the-few-step-video-inference-barrier-via-tr
- pare-pruning-and-adaptive-routing-for-efficient-video-genera
- spade-an-input-adaptive-sparse-attention-engine-for-fast-vid
- squad-sub-quadratic-attention-distillation-for-efficient-vid
- steering-video-diffusion-transformers-with-massive-activatio
- token-radius-attention-for-efficient-video-generation
Nano Banana 2
Mentioned in 7 entries
- aleph-20
- can-nano-banana-2-replace-traditional-image-restoration-mode
- gemini-can-now-pull-from-google-photos-to-generate-personali
- heres-this-weeks-shipping-recap-nano-banana-2-nano-banana-pr
- kimi-k26-nano-banana-2-pixel-perfect-images
- model-test-sensenova-u1-vs-gpt-image2-vs-nano-banana-in-info
- nano-banana-2-and-nano-banana-pro-are-generally-available-an
Opus 4.5
Mentioned in 7 entries
- big-gains-in-potatomonetbench-from-opus-45-to-fable-there-ar
- datasette-code-frequency-chart-on-github
- fable-5-is-the-biggest-step-up-ive-felt-in-our-models-since-
- it-is-going-to-be-like-what-happened-in-coding-as-soon-as-mo
- this-is-basically-an-opus-4546-level-model-that-will-work-on
- with-release-of-deepseek-v4-i-wanted-see-how-the-model-sizes
- xiami-mimo-v25-pro-mit-license-surpasses-opus-45-on-arena
CLAP
Mentioned in 7 entries
- comet-concept-space-dissection-of-the-modality-gap-in-audio-
- drift-augmented-scoring-text-derived-noise-robustness-for-ze
- figma-towards-fine-grained-music-retrieval
- membership-inference-for-contrastive-pre-training-models-wit
- phylogenetic-signal-in-marine-mammal-and-bird-vocalizations-
- probing-low-level-acoustic-attribute-encoding-in-clap-audio-
- the-sound-of-absence-audio-language-embedding-models-struggl
Nano Banana Pro
Mentioned in 7 entries
- beyond-generating-high-fidelity-visuals-we-wanted-to-test-th
- calling-all-builders-google-ai-pro-and-ultra-subscribers-wil
- heres-this-weeks-shipping-recap-nano-banana-2-nano-banana-pr
- ideogram4-vs-flux2-dev-vs-gpt-image-2-vs-nano-banana-pro
- nano-banana-2-and-nano-banana-pro-are-generally-available-an
- nano-banana-pro-vs-gpt-image-2-nano-banana-pro-wins-on-photo
- z-image-an-efficient-image-generation-foundation-model-with-
MAI-Thinking-1
Mentioned in 7 entries
- ainews-microsoft-build-mai-thinking-1-and-mai-family-models
- how-do-you-make-an-llm-anyway-microsoft-just-published-a-tex
- it-is-difficult-to-know-how-good-mai-thinking-1-is-from-the-
- microsoft-debuts-mai-thinking-1-its-first-advanced-reasoning
- microsofts-first-advanced-reasoning-ai-is-here
- microsofts-new-mai-models
- really-enjoyed-reading-the-microsoft-mai-thinking-1-building
Amazon Nova 2 Sonic
Mentioned in 7 entries
- build-a-healthcare-appointment-agent-with-amazon-nova-2-soni
- build-real-time-voice-streaming-applications-with-amazon-nov
- how-loka-built-a-natural-low-latency-voice-agent-with-amazon
- migrating-a-text-agent-to-a-voice-assistant-with-amazon-nova
- omnichannel-ordering-with-amazon-bedrock-agentcore-and-amazo
- real-time-voice-agents-with-stream-vision-agents-and-amazon-
- sciencesofts-hipaa-compliant-ai-voice-scheduler-built-on-aws
Decision Transformer
Mentioned in 7 entries
- decision-transformer-for-uav-mounted-ris-assisted-dynamic-d2
- e2dt-efficient-and-effective-decision-transformer-with-exper
- generalized-policy-gradient-with-history-aware-decision-tran
- goal-conditioned-decision-transformer-for-multi-goal-offline
- hybrid-offline-online-multi-agent-decision-transformers-for-
- trust-guided-decision-transformer
- wisdom-wireless-sparse-decision-transformer-with-mixture-of-
Nano Banana 2 Lite
Mentioned in 7 entries
- google-launches-nano-banana-2-lite-a-low-cost-text-to-image-
- googles-new-nano-banana-2-lite-image-model-is-its-fastest-an
- nano-banana-2-lite-gemini-31-flash-lite-image-now-on-ai-gate
- nano-banana-2-lite-image-edit-httpslinkscomfyorg4esb7gb
- nano-banana-2-lite-is-now-in-comfy-the-fastest-cheapest-mode
- nano-banana-2-lite-text-to-image-httpslinkscomfyorg3rdz92j
- start-building-with-nano-banana-2-lite-and-gemini-omni-flash
Llama 3
Mentioned in 7 entries
- built-gpt-2-llama-3-and-deepseek-from-scratch-in-pytorch---o
- fully-homomorphic-encryption-on-llama-3-model-for-privacy-pr
- how-we-optimized-a-local-llama-3-agent-from-15s-latency-and-
- k3-already-got-in-the-top-5-most-liked-models-of-all-time-on
- ollama-llama-3-in-a-small-python-app-modelfile-vs-long-promp
- ollama-run-llamar3-is-this-off-line
- testing-llama-3-locally-the-model-wasnt-the-issue-the-workfl
Video Diffusion Models
Mentioned in 7 entries
- core-co-evolving-reward-models-for-mitigating-latent-reward-
- dynamic-in-few-step-unifying-dynamic-computation-and-few-ste
- inversecrafter-efficient-video-recapture-as-a-latent-domain-
- latent-evolving-world-action-model
- navicache-test-time-self-calibration-caching-for-video-gener
- physics-aware-video-generation-via-agentic-planning-and-grap
- real2sam2real-generative-3d-caches-as-complementary-context-
GPT-Red
Mentioned in 7 entries
- ai-agents-are-already-being-used-to-improve-the-capabilities
- gpt-red-automated-red-teaming-via-self-play-at-scale
- introducing-gpt-red-an-internal-automated-red-teamer-on-a-mi
- meet-gpt-red-an-llm-super-hacker-openai-built-to-make-its-mo
- openai-details-gpt-red-an-ai-that-attacks-its-own-models-to-
- openai-gpt-red-self-improvement-for-robustness
- training-against-gptred-makes-gpt56-substantially-more-resil
Variational Autoencoders
Mentioned in 7 entries
- entropy-based-characterisation-of-the-polarised-regime-in-la
- generalization-in-vae-and-diffusion-models-a-unified-informa
- hyvic-a-metric-driven-spatio-spectral-hyperspectral-image-co
- improving-faint-object-detection-for-space-situational-aware
- p3-probabilistic-policy-propagation-for-stable-vae-based-rob
- unlocking-latent-dimensions-exploring-representations-of-lar
- yield-curves-dynamics-using-variational-autoencoders-under-n
Qwen 35B
Mentioned in 7 entries
- if-anyone-is-running-qwen-9b-or-27b-or-35b-and-getting-wrong
- increasing-active-parameters-per-token-in-moe-qwen-35b-a4b-r
- newer-commits-removed-the-qwen-35b
- nnoticing-qwen-27bq2-better-than-qwen-35bq8
- smallest-model-tips-for-intelligent-computer-use-via-hermes
- training-free-hack-to-reduce-moe-reasoning-tokens-by-85-with
- yandexaliceai-foundation-80b-a3b-base-russian-developed-comp
MLP
Mentioned in 7 entries
- a-comparative-analysis-of-mlp-and-kolmogorov-arnold-networks
- a-multimodal-label-forecasting-method-for-aperiodic-visuo-mo
- benchmarking-optimizers-for-mlps-in-tabular-deep-learning
- delta-position-estimation-based-imu-odometry-a-comparison-of
- kolmogorov-arnold-networks-kan-for-aerodynamic-prediction-a-
- tabpack-efficient-hyperparameter-ensembles-for-tabular-deep-
- timepre-bridging-accuracy-efficiency-and-stability-in-probab
SigLIP
Mentioned in 7 entries
- b11043
- gate-reliability-gated-gaussian-evidence-fusion-for-training
- koda-contrastive-representation-comparison-and-alignment-for
- oops-sigreg-did-it-again-large-scale-cc12m-datacomp-l-vision
- we-introduce-levljepa-the-first-fully-non-contrastive-end-to
- weather-robust-scene-semantics-with-vision-aligned-4d-radar
- zooclaw-fashionsiglip2-distilled-fine-tuning-for-robust-fash
Omni
Mentioned in 7 entries
- ainews-google-io-2026-gemini-35-flash-omni-nanobanana-for-vi
- context-unrolling-in-omni-models
- google-just-revealed-omni-personalized-cross-device-intellig
- google-launches-new-omni-model-google-says-that-today-omni-c
- rowancheung-anchored-our-podcast-porch-and-interviewed-sunda
- thinking-machines-just-dropped-a-1t-omni-model-1m-context-wi
- watch-your-avatar-speak-spanish-english-and-japanese-httpsxc
MiniMax
Mentioned in 7 entries
- anyone-wonder-why-americans-lag-so-far-behind-in-the-open-ll
- best-minimax-configuration
- introducing-claude-for-music-you-can-now-create-songs-from-c
- ive-tried-cursor-claude-code-ollama-etc-but-i-still-dont-kno
- like-if-you-have-been-using-qwen-kimi-minimax-it-feels-like-
- run-minimax-models-on-amazon-bedrock
- yst-on-stream-codex-wrote-a-skill-for-my-hermes-agent-to-mak
Mythos Preview
Mentioned in 7 entries
- anthropic-researchers-say-mythos-preview-can-now-turn-public
- gpt-55-matches-heavily-hyped-mythos-preview-in-new-cybersecu
- gpt-56-sol-matches-mythos-preview-on-exploitbench-adds-ultra
- mythos-preview-is-the-first-ai-model-to-complete-both-of-ais
- sources-cisa-doesnt-have-access-to-mythos-preview-even-as-so
- sources-say-nsa-is-using-mythos-preview-and-a-source-says-it
- the-uk-aisi-found-mythos-preview-is-the-first-model-to-solve
Mythos 5.1
Mentioned in 7 entries
- anthropic-gives-eu-cybersecurity-agency-enisa-testing-access
- anthropic-launches-claude-fable-51-and-says-it8217s-up-to-45
- anthropic-unveils-claude-fable-51-which-is-generally-availab
- claude-fable-51-and-mythos-51-are-anthropics-first-models-to
- lwiai-podcast-256---fable-51-astra-tease-gemini-38-flash
- sources-anthropic-declined-to-submit-mythos-51-to-the-uk-ais
- we-really-need-to-get-someone-experienced-in-ai-with-strong-
Gemini 3.6 Flash
Mentioned in 7 entries
- deepseek-v4-flash-scores-50-on-the-artificial-analysis-intel
- google-launches-a-cheaper-alternative-to-large-ai-security-m
- google-says-free-gemini-users-will-be-limited-to-the-35-flas
- introducing-gemini-36-flash-35-flash-lite-and-35-flash-cyber
- llm-gemini-033
- openwiki-now-supports-gemini-ai-studio-enterprise-vertex-ai-
- we-benchmarked-gemini-36-flash-and-gemini-35-flash-lite-on-d
deepseek-v4-flash
Mentioned in 7 entries
- api-is-available-today-keep-baseurl-just-update-model-to-dee
- b10549
- cant-find-deepseek-v4-flash-after-adding-api-key-ollama-pro
- deepseek-v4-flash-is-now-available-on-ollamas-cloud-hosted-i
- httpsxcomollamastatus2047598971435290992s20
- i-subscribed-to-ollama-pro-today-deepseek-v4-flash0731-works
- nice-stats-on-usage-of-open-models-across-opencode-glm-52-is
Deepseek
Mentioned in 7 entries
- as-president-trump-meets-president-xi-this-week-a-call-to-th
- best-40b-alternatives-to-qwendeepseek-for-1-coding-2-long-do
- deepseek-ollama-openclaw-fully-local-0-heres-what-you-actual
- do-you-know-you-can-use-open-weight-models-like-glm-deepseek
- ollama-cloud-with-deepseek-rate-limit
- our-gb300-cluster-went-down-yesterday-just-as-deepseek-relea
- rampart-our-pii-removal-model-has-cracked-the-first-screen-o
Qwen3.5 4B
Mentioned in 7 entries
- a-jev-style-model-fine-tuned-on-qwen35-4b
- can-you-use-qwen35-4b-gemma-4-e4b-with-z-imageturbo
- how-to-train-your-own-jev-for-17
- is-there-a-better-small-model-than-qwen35-4b-for-a-fast-loca
- procedural-skill-sft-across-capacity-tiers-a-w-shaped-pre-sf
- qwen35-4b-grabbing-logits-is-almost-jev-or-even-just-qwen-re
- were-releasing-tev1-4b-experimental-a-jev-like-classifier-fi
Qwen Image 2.1
Mentioned in 7 entries
- custom-node-for-extract-prompt-target-aspect-ratiodimensions
- deterministic-kittens-fun-with-qwen-image-21-on-an-m2-macboo
- fizgig-650---qwen-image-21-and-a-new-qwen-image-21-training-
- qwen-image-21-generation-fix-lora-and-settings
- qwen-image-21-masked-inpainting-working-crop-and-stitch-grap
- qwen-image-21-releasing-tomorrow
- train-slider-loras-for-qwen-image-21-in-fizgig-670
Gemini 4 Argon
Mentioned in 7 entries
- artificial-analysis-says-gemini-4-argon-high-matches-gpt-6-a
- google-announces-gemini-4-and-says-it8217s-so-capable-that-o
- google-rolls-out-gemini-4-argon-to-a-small-group-of-cybersec
- googles-new-frontier-ai-model-gemini-4-argon-goes-to-cyberse
- learn-more-httpsbloggoogleinnovation-and-aimodels-and-resear
- lots-of-discussion-out-there-about-our-next-model-so-i-wante
- timelines-just-got-shorter-ai-will-understand-the-physical-3
FLUX.2 Klein
Mentioned in 6 entries
- any-good-settings-for-character-lora-for-klein
- color-anchor-node-flux2klein
- flux-klein-9b-pixel-space---comfyui-nodes
- flux2-klein-2-stage-upscale
- multi-angle-lora-for-flux-klein
- using-klein-enhancer-node-for-anime-to-real
MeanFlow
Mentioned in 6 entries
- extending-one-step-image-generation-from-class-labels-to-tex
- flow-map-grpo-reinforcement-learning-for-few-step-flow-map-g
- flowmap-opd-rollout--kernel-separation-for-on-policy-distill
- improved-mean-flows-on-the-challenges-of-fastforward-generat
- large-scale-diffusion-distillation-via-score-regularized-con
- meno-meanflow-enhanced-neural-operators-for-dynamical-system
Qwen3.6-Plus
Mentioned in 6 entries
- introducing-qwen36-max-preview-an-early-preview-of-our-next-
- introducing-qwen36-plus-from-alibabaqwen-a-1m-context-model-
- its-finally-here-huge-thanks-to-the-opencode-team-for-the-se
- qwen36-plus-is-now-available-on-together-ai-try-it-now-httpw
- qwen36-plus-just-swept-openrouters-daily-weekly-trending-cha
- sub-agent-model-selection-different-tasks-different-models-y
Qwen 3.6 Plus
Mentioned in 6 entries
- 3million-output-tokens-qwen-35-plus-is-basically-a-frontier-
- alibabaqwen-sign-up-to-get-free-access-to-qwen-36-plus-and-m
- qwen-36-plus-by-alibabaqwen-is-now-free-for-a-limited-time-o
- qwen-36-plus-on-ai-gateway
- qwen36-plus-is-on-nous-portal-now-and-free-for-a-limited-tim
- the-future-of-ai-is-agent-native-excited-to-kick-off-this-jo
Gemini 3.1
Mentioned in 6 entries
- glm-51-by-zaiorg-is-now-3-in-code-arena---surpassing-gemini-
- google-home-gets-upgraded-gemini-voice-assistant-and-new-cam
- google-home8217s-gemini-ai-can-handle-more-complicated-reque
- i-think-artificial-analysis-does-a-good-job-overall-and-prov
- process-speech-as-speech-with-gpt-realtime-2-or-gemini-31-li
- sparks-unicorn-httpsxcomemollickstatus2024756029121020236s20
LTX-2.3-22b
Mentioned in 6 entries
- editanything-ic-lora---ltx-23
- httpshuggingfacecolightricksltx-23-22b-ic-lora-cross-eyed-al
- lipdub-workflow-ltx-23-ic-lora
- ltx-23-union-control-sdxl
- ltx23---ltx-23-22b-ic-lora-outpaint
- vr-outpaint-ic-lora-for-ltx23-released
Ernie Image Turbo
Mentioned in 6 entries
- amazing-style-made-using-ernie-image-turbo
- ernie-image-turbo---i-like-it-but-the-bias-is-too-strong
- ernie-image-turbo-excellent-model-without-any-blonde
- ernie-image-turbo-is-not-bad-at-all-using-int8-quant-and-gem
- try-this-prompt-in-flux-2-klein-9b-ernie-image-turbo-and-z-i
- turns-out-ernie-image-turbo-is-quite-well-versed-in-anime
Chroma
Mentioned in 6 entries
- 2-the-same-message-applies-to-several-models-chroma-z-image-
- 3-the-same-message-applies-to-several-models-chroma-z-image-
- 4-a-mesma-mensagem-aplica-se-a-vrios-modelos-chroma-z-image-
- 5-the-same-message-applies-to-several-models-chroma-z-image-
- same-prompt-for-various-models---chroma-z-image-klein-qwen-e
- tier-list-between-fluxqwenanimachromaz-image
Klein 9b
Mentioned in 6 entries
- anime-2-real-with-flux-2-klein-9b
- ideogram-40-vs-zib-vs-klein-9b
- teal-dark---flux2-klein-9b-styleaesthetic-lora
- the-essential-calvin-hobbes---flux2-klein-9b-base---4x-upsca
- weird-colorful-pattern-for-klein-9b-model
- working-on-a-technique-to-produce-style-loras-from-a-single-
Cosmos 3
Mentioned in 6 entries
- ainews-nvidia-cosmos-3-nemotron-3-ultra-and-rtx-spark
- cosmos-3-omnimodal-world-models-for-physical-ai
- how-cosmos-3-helps-physical-ai-think-before-it-acts
- microsoft-nvidia-roundup-of-links-httpswwwlatentspacepainews
- nvidia-releasescosmos3-super-text2image-model-64-billion-par
- nvidias-new-ai-changed-robotics-forever
SenseNova U1
Mentioned in 6 entries
- github-comfyui-sensenova-u1-released-anyone-got-it-working-y
- model-test-sensenova-u1-vs-gpt-image2-vs-nano-banana-in-info
- sensenova-u1-comfyui-node-8-step-lora-support-and-gguf-vramr
- sensenova-u1-infographic-test-capabilities-in-image-based-re
- sensenova-u1-infographic-test-high-text-fidelity-even-in-inf
- sensenova-u1-with-neo-unify-just-dropped
Qwen2.5-VL
Mentioned in 6 entries
- b9453
- best-local-vision-language-models
- how-good-ollama-qwen25-vl-for-dataectraction-from-pdf-like-o
- i-built-a-local-qwen25-vl-desktop-tool-that-lets-you-ask-que
- sepprunea-separator-based-pruning-framework-for-efficient-mu
- tuning-qwen25-vl-to-improve-its-web-interaction-skills
Spiking Neural Networks
Mentioned in 6 entries
- badsnn-backdoor-attacks-on-spiking-neural-networks-via-adver
- neuromorphic-parameter-estimation-for-power-converter-health
- not-all-timesteps-matter-equally-selective-alignment-knowled
- spiker-ll-an-energy-efficient-fpga-accelerator-enabling-adap
- unsupervised-spiking-feature-learning-for-event-based-pedest
- vmem-arphi-low-compute-out-of-distribution-detection-in-spik
LeWorldModel
Mentioned in 6 entries
- fast-leworldmodel
- flow-jepa-flow-matching-for-robust-latent-dynamics-in-jepa-w
- mind-the-gap-promises-and-pitfalls-of-hierarchical-planning-
- sub-jepa-a-simple-fix-to-lecun-groups-leworldmodel-that-cons
- the-evaluation-protocol-determines-the-result-an-independent
- yann-lecun-closed-103b-for-ami-labs-on-march-10-three-days-l
Krea 2 Turbo
Mentioned in 6 entries
- krea-2-is-back-and-this-time-the-weights-are-open-this-open-
- krea2-turbo-distill-4-step-lora---new-checkpoint-chk60k-rele
- qwen-21-styles-comprehension
- svdquant-native-int8w4a4-for-krea-2-on-comfyui-up-to-2x-fast
- theres-no-one-weird-trick-for-prompting-krea-2-art-stylesjus
- today-we-release-the-open-weights-of-krea-2-welcome-krea-2-r
YOLO26
Mentioned in 6 entries
- dental-panoramic-radiograph-analysis-using-yolo26-from-tooth
- does-yolo26-truly-offer-advantages-over-its-predecessors-for
- mambapsa-a-mamba-based-replacement-for-c2psa-in-yolo26
- multiscale-real-time-object-detection-in-the-nms-free-era-a-
- toward-optimal-adenovirus-detection-using-yolo26
- ultralytics-yolo-evolution-an-overview-of-yolo27-yolo26-yolo
Tsetlin Machine
Mentioned in 6 entries
- a-real-time-tsetlin-machine-based-non-intrusive-load-monitor
- autonomous-collaborative-learning-among-an-ensemble-of-tsetl
- compressed-recurrent-feedback-in-tsetlin-machines-a-reproduc
- leveraging-interpretable-tsetlin-machine-for-pdf-malware-det
- on-device-interpretable-tsetlin-machine-based-intrusion-dete
- the-tsetlin-machine-goes-deep-logical-learning-and-reasoning
DeepSeek V3
Mentioned in 6 entries
- accelerating-dropless-moe-training-in-jax-with-nvidia-transf
- deepseek-founders-4-hour-investor-meeting-deepseek-is-priori
- deepseek-training-2t-and-plans-8t-model
- deepseeks-new-ai-is-a-game-changer
- did-they-shut-down-deep-seek-cloud-for-free-users
- just-a-reminder-that-deepseek-v3-came-out-18-months-ago-and-
GPT 5.6
Mentioned in 6 entries
- annoying-that-openai-doesnt-seem-to-give-a-gdpval-measure-fo
- garys-analysis-is-pretty-much-spot-on-as-ever-this-is-such-a
- openai-released-gpt-56-today-and-we-ran-a-day-0-benchmark-in
- people-are-not-getting-how-significant-this-news-the-white-h
- so-on-brand
- sources-sam-altman-told-staff-the-us-government-asked-openai
xLSTM
Mentioned in 6 entries
- a-deep-reinforcement-learning-approach-to-automated-stock-tr
- aga3dnet-anatomy-guided-gaussian-priors-with-multi-view-xlst
- benchmarking-transformer-and-xlstm-for-time-series-forecasti
- effective-distillation-to-hybrid-xlstm-architectures
- tirex-2-generalizing-tirex-to-multivariate-data-and-streamin
- x-track-physics-aware-xlstm-for-realistic-vehicle-trajectory
DLSS 5
Mentioned in 6 entries
- how-do-i-replicate-the-dlss-5-output-using-flux-klein
- i-built-a-standalone-dlss-5-neural-rendering-video-tool-no-r
- nba-2k27-with-nvidia-dlss-5-leads-28-new-games-coming-to-gef
- open-sourced-an-experimental-standalone-dlss-5-video-player-
- title-mlx-dlss-nvidias-dlss-5-neural-renderer-and-frame-gene
- whats-new-for-game-developers-dlss-5-with-3d-guided-neural-r
Laguna M.1
Mentioned in 6 entries
- add-laguna-m1-gguf-support-by-empty-quiver-pull-request-2003
- laguna-m1xs2-technical-report
- loving-the-latentspacepod-breakdown-of-our-laguna-m1xs2-tech
- today-poolsideai-is-releasing-laguna-m1-laguna-xs2-our-lates
- today-were-shipping-laguna-m1-and-laguna-xs2-our-first-publi
- us-startup-poolside-debuts-its-first-open-weight-model-lagun
Fourier Neural Operator
Mentioned in 6 entries
- conformal-prediction-for-neural-operators-distribution-free-
- fourier-neural-operators-for-rayleigh-benard-convection
- higher-order-fourier-neural-operator-explicit-mode-mixer-for
- hs-fno-history-space-fourier-neural-operator-for-non-markovi
- mitigating-spectral-bias-in-neural-operators-for-underwater-
- physics-informed-conformal-prediction-embedding-pde-consiste
Muon
Mentioned in 6 entries
- denoise-first-orthogonalize-later-understanding-momentum-in-
- efficient-long-horizon-learning-for-learned-optimization
- htmuon-improving-muon-via-heavy-tailed-spectral-correction
- ringmaster-lmo-asynchronous-linear-minimization-oracle-momen
- sign-compression-for-muon-signmuon-muonsign-and-the-limits-o
- spectral-gradient-descent-mitigates-anisotropy-driven-misali
T5
Mentioned in 6 entries
- from-bert-to-t5-a-study-of-named-entity-recognition
- investigating-temporal-motion-features-for-pose-to-text-indi
- large-language-models-are-universal-reasoners-for-visual-gen
- most-transformer-modifications-still-do-not-transfer-at-1-3b
- sentence-splitter-uncovering-latent-factual-structure-for-se
- the-maskability-index-predicting-task-objective-alignment-in
ChatGPT 5.5
Mentioned in 6 entries
- adult-mode-on-55
- blown-away-by-chatgpt-55-image-gen-paris-in-1995
- cant-agree-more-after-trying-out-chatgpt-5455-for-two-months
- chatgpt-55-2026-talks-to-cleverbot-2006
- chatgpt-55-is-here
- did-anyone-suddenly-lose-access-to-54-and-55
Flow Matching
Mentioned in 6 entries
- a-lagrangian-view-of-flow-matching
- alleviating-sparse-rewards-by-modeling-step-wise-and-long-te
- drift-flow-matching
- flow-opd-on-policy-distillation-for-flow-matching-models
- geoflow-efficient-driving-video-generation-via-geometry-alig
- online-evolution-strategy-for-flow-matching-vla-policies-via
Qwen3-8B
Mentioned in 6 entries
- benchmarking-linguistic-adaptation-in-comparable-sized-llms-
- build-an-ai-powered-product-tagging-system-with-amazon-sagem
- fine-tuning-general-purpose-large-language-models-for-agricu
- impressive-paper-showing-the-impact-of-a-good-harness-improv
- linguiutics-at-psydefdetect-iterative-imbalance-aware-fine-t
- scaling-post-training-ternarisation-to-qwen3-8b-capability-r
AlphaFold3
Mentioned in 6 entries
- a-newly-released-ai-tool-has-generated-an-atlas-of-more-than
- agentic-baim-llm-evaluation-able-benchmarking-llm-use-of-pro
- dcfold-efficient-protein-structure-generation-with-single-fo
- esmfold2-is-a-state-of-the-art-folding-model-its-crazy-good-
- inference-time-optimization-for-experiment-grounded-protein-
- probing-and-steering-biology-across-boltz-1s-trunk-diffusion
RoBERTa
Mentioned in 6 entries
- contextual-trajectory-and-incremental-contextual-displacemen
- hallelubert-let-every-token-that-has-meaning-bear-its-weight
- polish-modernbert-the-long-and-short-of-polish-language-unde
- sg-unibuc-nlp-at-semeval-2026-task-6-multi-head-roberta-with
- transformer-models-for-text-summarization-a-comparative-stud
- ucsc-nlp-at-semeval-2026-task-10-boundary-aware-span-extract
SWE-1.7
Mentioned in 6 entries
- cognition-releases-swe-17-trained-from-kimi-k27-and-availabl
- for-coding-agents-trustworthiness-has-to-be-tested-in-the-ha
- introducing-swe-17-the-most-capable-model-weve-trained-yet-i
- swe-17-is-available-today-in-devin-across-web-desktop-and-cl
- swe-17-is-free-for-all-paid-users-for-the-next-month-were-al
- yesterday-we-launched-swe-17-built-on-the-open-source-kimi-k
Qwen3.6-35B
Mentioned in 6 entries
- i-forked-ninfer-3090-and-converted-it-to-run-on-the-cmp170hx
- nanbeige42-3b-im-not-impressed
- new-benchmark-just-dropped
- qwen36-35b-becomes-competitive-with-cloud-models-when-paired
- searching-for-38-35b-qwen36-35b-a3b-testing-5-finetunes-vs-b
- wait-what-2-bit-qwen36-35b-a3b-is-lightning-fast-and-it-only
Hy3
Mentioned in 6 entries
- deepseek-v4-flash---hy3-or-is-qwen36-27b-still-the-most-soli
- hy3-the-new-295b-moe-model-from-tencenthunyuan-is-now-free-i
- new-model-for-amd-strix-halo-users-my-198b-step-37-flash-rel
- other-active-promotions---free-models-solar-pro-4-1-week-hy3
- tencenthy3
- two-weeks-after-release-hy3-preview-is-1-on-openrouters-week
DeepSeek V3.2
Mentioned in 6 entries
- b10237
- b11020
- deepseek-v4---almost-on-the-frontier-a-fraction-of-the-price
- do-not-blindly-delete-your-older-models-some-are-still-preci
- how-we-built-the-most-performant-deepseek-v32-minimax-m25-an
- need-info-on-quality-benchmarks-to-run-on-deepseek-v32-diffe
Kimi-K2.6
Mentioned in 6 entries
- kimi-k26-is-on-huggingface
- kimimoonshot-kimi-k26-just-dropped-in-qoder-sota-coding-long
- lwiai-podcast-242---chatgpt-images-20-qwen-36-max-kimi-k26
- moonshot-ai-releases-kimi-k26-model-with-1t-parameters-atten
- we-trained-and-released-dspark-speculators-for-kimi-k26-and-
- weve-quantized-kimi-k26-to-mxfp4-on-amd-download-and-use-tod
Meta AI
Mentioned in 6 entries
- in-a-notice-to-maines-ag-meta-says-20k-instagram-accounts-ma
- italys-competition-authority-says-it-has-dropped-its-2025-pr
- meta-is-alerting-instagram-users-whose-accounts-were-taken-o
- meta-is-making-its-ai-chatbot-more-like-an-assistant
- meta-rolls-out-plus-plans-for-instagram-facebook-and-whatsap
- sensor-tower-muse-was-downloaded-902k-times-in-the-six-days-
Aleph 2.0
Mentioned in 6 entries
- aleph-20-is-now-available-via-the-runway-api-bring-precise-v
- aleph-20-will-blow-your-mind
- not-so-fast-google-omni---runwayml-is-back-in-the-game-with-
- one-video-now-made-for-every-feed-and-format-upload-your-exi
- we-built-aleph-as-one-of-the-first-generalist-text-based-edi
- yesterday-we-released-aleph-20-our-upgraded-video-editing-mo
Claude Opus 4
Mentioned in 6 entries
- anthropic-blames-dystopian-sci-fi-for-training-ai-models-to-
- cool-that-grok-45-is-1-in-some-respects-even-with-respect-to
- how-close-are-models-to-building-a-perfect-slack-clone-with-
- i-upgraded-my-claude-token-counter-tool-to-compare-different
- qwen-38-27b-aider-score
- we-integrated-frontiercs-into-harbor-and-are-releasing-a-pre
GPT-4.1
Mentioned in 6 entries
- cross-model-consistency-of-ai-generated-exercise-prescriptio
- decomposing-wrong-consensus-agreement-in-llm-self-consistenc
- evaluating-the-effectiveness-of-persona-simulation-in-opinio
- llm-predictive-scoring-and-validation-inferring-experience-r
- what-would-it-take-for-the-frontier-labs-to-open-the-weights
- whats-new-in-microsoft-foundry-fine-tuning-april-2026
Gemini Flash
Mentioned in 6 entries
- 1-on-memory-benchmark-longmemeval-with-gemini-flash-not-pro-
- evaluating-advanced-prompting-on-gemini-flash-for-multi-hop-
- gemma-4-adoption-numbers-outpacing-qwen-3536-for-the-same-si
- help-odysseus-not-using-gemini-flash-for-images-deepseek-wri
- i-am-pleased-to-see-that-openais-new-model-is-roughly-on-par
- trained-my-first-small-language-model
DeepSeek Flash
Mentioned in 6 entries
- ainews-deepseek-v4-pro-16t-a49b-and-flash-284b-a13b-base-and
- another-harness-matters-post-codex-cli-pi-and-opencode
- congrats-to-deepseekai-team-doing-the-numbers-i-would-estima
- qwen-38-27b-vs-deepseek-flash
- something-has-happened-with-post-training-as-shown-by-deepse
- the-backgrounds-were-generated-with-minimax-m3-everything-el
Qwen 3.8 Max
Mentioned in 6 entries
- appreciate-it-high-performance-low-cost-try-it-out-today
- from-22-to-4-on-legal-research-bench-solid-progress-for-qwen
- is-glm-52-kimi-27-still-worth-it
- live-in-command-code
- mac-studio-m5-max-cost-analysis
- qwen-38-max-and-minimax-h3-within-hours-of-each-other
Claude Mythos 5
Mentioned in 6 entries
- anthropic-details-unreleased-model-2-new-alignment-concerns-
- anthropic-sets-ai-performance-records-with-new-mythos-5-fabl
- claude-fable-5-and-claude-mythos-5
- claude-mythos-5-built-a-malicious-python-package-created-acc
- openai-introduces-gpt-56-to-challenge-claude-mythos-5
- quoting-claude-opus-5-system-prompt
Nemotron 3
Mentioned in 6 entries
- b11205
- fine-tune-nvidia-nemotron-3-models-with-amazon-sagemaker-ai-
- microsoft-nvidia-roundup-of-links-httpswwwlatentspacepainews
- nvidia-launches-nemotron-3-nano-omni-an-open-multimodal-mode
- really-excited-to-see-this-release-and-integrate-it-with-dee
- rl-fine-tuning-is-now-live-for-nvidiaai-nemotron-3-on-firewo
Qwen 4
Mentioned in 6 entries
- alibaba-releases-qwen38-flash-an-open-weight-125b-parameter-
- guys-i-am-absolutely-astounded-the-qwen-36-27b-is-like-a-jum
- is-all-the-work-thats-being-put-into-qwen38-flash-next-going
- qwen-38-flash-next-is-based-on-qwen-4-architecture-if-the-an
- qwen-4-architecture-what-do-we-know
- qwen-will-be-the-king
GPT-6-Astra
Mentioned in 6 entries
- gpt-6-astra-in-a-navigation-workflow-behavioral-analysis-in-
- gpt-6-astra-lights-up-embodied-navigation-evaluation-in-zero
- openai-quietly-altered-the-performance-metrics-it-reported-f
- pause-openai---they-are-a-deeply-unethical-company-building-
- robot-manipulation-with-gpt-6-astra-body-knowledge-experienc
- seeing-how-differently-people-prompt-llms-is-funny
System One
Mentioned in 6 entries
- jev-introduces-a-new-shape-of-llm---system-one-aka-decision-
- own-your-harness-folks-i-feel-like-system-one-models-take-cu
- pay-attention-to-system-one-models-if-you-are-building-custo
- qwen35-4b-grabbing-logits-is-almost-jev-or-even-just-qwen-re
- this-is-brilliant-recommended-reading-its-a-great-example-of
- von-open-source-395m-system-one-model
Qwen 2.1
Mentioned in 6 entries
- 192-generated-image-side-by-side-qwen-image-21-vs-krea-2
- is-it-just-me-or-does-qwen-21-look-like-a-heavily-distilled-
- just-bought-a-second-3090-but-now-i-dont-see-the-benefits-ri
- qwen-21-detail-enhancer-lora
- qwen-21-is-a-major-disappointment-for-t2i
- qwen-21-styles-comprehension
Claude Sonnet 5.5
Mentioned in 6 entries
- anthropic-debuts-claude-sonnet-55-running-30-faster-than-the
- claude-sonnet-55
- claude-sonnet-55-is-live-in-hermes-agent-enjoy
- introducing-claude-sonnet-55-on-aws
- openai-prices-gpt-61-sol-at-21m-input-and-101m-output-tokens
- qwen-next-38-and-38-27b-vs-sonnet-55-low-and-sonnet-55-mediu
SAM3D
Mentioned in 5 entries
- fast-sam3d-3dfy-anything-in-images-but-faster
- mv-sam3d-adaptive-multi-view-fusion-for-layout-aware-3d-gene
- object-pose-and-shape-estimation-for-grasping-does-it-work
- sam3d-guided-object-centric-representation-alignment-for-vis
- sam3dual-a-3rd-place-solution-to-the-mosev2-track-8th-lsvos-
FSD Supervised
Mentioned in 5 entries
- for-a-driver-born-without-arms-fsd-supervised-is-life-changi
- fsd-supervised-approved-in-the-netherlands
- good-news-for-belgian-tesla-owners-the-netherlands-rdw-just-
- tesla-insurance-update-with-the-latest-version-of-safety-sco
- tesla-owners-in-the-netherlands-have-driven-10-million-km-on
LVLMs
Mentioned in 5 entries
- countering-the-over-reliance-trap-mitigating-object-hallucin
- distilling-visual-reasoning-into-text-space
- persistent-visual-memory-sustaining-perception-for-deep-gene
- visreflect-latent-visual-reflection-for-fine-grained-percept
- when-does-a-video-language-model-stop-watching-reward-streng
Haiku
Mentioned in 5 entries
- anthropic-and-openai-will-increasingly-be-able-to-serve-weak
- just-shipped-advisor-middleware-an-open-source-implementatio
- open-models-a-potential-looming-haiku-apocalypse-chart-compa
- the-labs-should-launch-some-easy-to-view-comparisons-of-thei
- when-boomer-companies-get-high-anthropic-bill-they-set-spend
Physics-Informed Neural Networks
Mentioned in 5 entries
- depth-resolved-coral-reef-thermal-fields-from-satellite-sst-
- feature-interaction-modeling-for-physics-informed-neural-net
- label-propagation-for-physics-informed-neural-networks-and-p
- multi-scale-dynamic-wake-modeling-of-floating-offshore-wind-
- physics-informed-neural-networks-for-joint-source-and-parame
MLLM
Mentioned in 5 entries
- gameworld-towards-standardized-and-verifiable-evaluation-of-
- icg-improving-cover-image-generation-via-mllm-based-promptin
- sa2va-marrying-sam2-with-mllm-for-dense-grounded-understandi
- spar3d-moe-adaptive-3d-spatial-reasoning-from-sparse-views-m
- video-rate-streaming-stylization-on-a-vision-aware-mllm-cond
ERNIE Image Turbo
Mentioned in 5 entries
- a-new-way-to-reduce-the-grid-on-ernie-image-turbo
- ernie-and-a-complex-composition-in-one-run-guest-zit-details
- ernie-image-turbo-is-capable-of
- i-have-extracted-the-lora-from-ernie-image-turbo
- i-tested-ernie-image-turbo-fp8-nvfp4-fp16-and-int8-with-nano
IC-LoRA
Mentioned in 5 entries
- a-look-at-the-video-outpainting-workflow-in-comfyui-using-lt
- bodypositivity-ic-lora-for-ltx23-is-out-now
- dramabox---most-expressive-voice-model-ever-based-on-ltx-23
- ic-loras-for-ltx2-have-so-much-potential---you-can-train-sot
- lipdub-beta-new-open-source-lipsync-ic-lora
ERNIE Image
Mentioned in 5 entries
- ernie-image-ernie-turbo-lora-elusarcas-anime-style
- ernie-image-nvfp4-workflow-optional-turbo-lora-prompt-enhanc
- ernie-image-released
- more-details-httpsblogcomfyorgpernie-image-day-0-support
- testing-ideogram-json-prompts-in-ernie-image
Flux.2 Dev
Mentioned in 5 entries
- character-workflow-chroma1-hd-flux2-dev-wan-22-ltx-23
- chroma1-hd-character-transfer-with-flux2-dev
- comparison-of-low-steps-klein-9b-x-z-image-turbo-x-ernie-tur
- flux2klein-little-info
- ideogram4-vs-flux2-dev-vs-gpt-image-2-vs-nano-banana-pro
Flux.2 Klein 9B
Mentioned in 5 entries
- comparison-of-low-steps-klein-9b-x-z-image-turbo-x-ernie-tur
- flux2-klein-9b-4b-scribbly-doodle-lora
- flux2-klein-9b-lcs-consistency-lora-20260415---maximum-color
- how-to-fix-extra-limbs-in-flux2-klein-9b
- psionix-90s-comic-lora-for-flux2-klein-9b
LFM2
Mentioned in 5 entries
- b10549
- b11190
- lfm25-encoders-fast-at-long-context-even-on-cpu
- liquidailfm25-encoder-250m350m
- tests-of-the-hybrid-models
Grok Voice Think Fast 1.0
Mentioned in 5 entries
- grok-voice-think-fast-10-is-officially-the-most-well-rounded
- grok-voice-think-fast-10-set-the-bar-1-spot-a-strong-start-6
- think-fast
- try-grok-voice
- try-grok-voice-for-your-customer-support
Kimi 2.6
Mentioned in 5 entries
- deepseek-v4-works-fine-but-its-not-the-frontier-pressing-mom
- llm-openrouter-06
- opencode-with-kimi-26-via-ollama-cloud-said-thinking--35-tur
- so-excited-for-this---in-the-past-four-months-ive-written-al
- thank-you-to-kimimoonshot-for-adding-an-additional-1000-to-t
Z image
Mentioned in 5 entries
- 2-the-same-message-applies-to-several-models-chroma-z-image-
- 3-the-same-message-applies-to-several-models-chroma-z-image-
- 4-a-mesma-mensagem-aplica-se-a-vrios-modelos-chroma-z-image-
- 5-the-same-message-applies-to-several-models-chroma-z-image-
- same-prompt-for-various-models---chroma-z-image-klein-qwen-e
Qwen3.5-397B-A17B
Mentioned in 5 entries
- alibaba-launches-qwen36-27b-an-open-weight-dense-model-with-
- lm-performancewith-only-27b-parameters-qwen36-27b-outperform
- run-qwen36-27b-locally-unslothai
- running-qwen35-397b-a17b-4bit-quants-177-gb-on-two-dgx-spark
- thanks-to-lmsysorg-try-it-on-sglang-now
Gemma-4
Mentioned in 5 entries
- b10246
- b10981
- b8860
- codex-is-overoptimised-for-large-models-it-ranks-2nd-out-of-
- trials-and-tribulations-fine-tuning-deploying-gemma-4-p
ERNIE
Mentioned in 5 entries
- ernie---only-asians
- ernie-image-lora-training---my-take
- great-news-the-ernie-editing-model-is-expected-to-be-release
- i-like-that-there-are-models-called-bert-and-ernie-but-in-al
- text-encoder-ablitered-prompt-enchanced-ablitered---for-erni
PixelDiT
Mentioned in 5 entries
- comfyui-v0230-support-nvidia-pixeldit-and-pid-core-201-by-ki
- pixeldit-13b-pixel-space-diffusion-transformer-no-vae-4gb-vr
- pixeldit-comfyui-wen
- pixeldit-pixel-diffusion-transformers-for-image-generation
- pixeldit-pixel-diffusion-transformers-for-image-generation-p
Flux.2
Mentioned in 5 entries
- amuse-ai-flux2-on-9060xt-16gb
- blind-realism-test-z-image-turbo-vs-klein-9b-distilled
- flux2-klein-pipeline-for-real-time-webcam-stream-processing-
- flux2-klein-spectral-graft---a-node-for-addingremoving-objec
- teal-dark---flux2-klein-9b-styleaesthetic-lora
MiMo-V2.5
Mentioned in 5 entries
- b10155
- full-pipeline-inference-optimization-for-mimo-v25-series-pus
- mimo-v25-gguf-preview-available
- perky-little-fella-from-mimo-v25
- xiaomis-mimo-v25-and-mimo-v25-pro-are-both-now-available-in-
OpenVLA
Mentioned in 5 entries
- crossvla-cross-paradigm-post-training-and-inference-optimiza
- early-warning-signals-for-openvla-failure-under-visual-distr
- looking-for-real-world-comparisons-between-wall-oss-pi06-and
- refactor-vla-unsupervised-library-learning-of-typed-motor-pr
- reinforcement-fine-tuning-of-flow-matching-policies-for-visi
Claude Pro
Mentioned in 5 entries
- anyone-else-able-to-duplicate-the-34month-pricing-for-claude
- claude-pro-feels-much-more-dynamic-and-visual-than-chatgpt-p
- i-actually-cancelled-my-claude-max-subscription-well-downgra
- one-of-the-easiest-ways-to-show-journalists-and-academics-wh
- ran-chatgpt-plus-and-claude-pro-side-by-side-for-30-days-her
Gaussian process
Mentioned in 5 entries
- improving-bayesian-optimization-via-training-aware-condition
- information-based-calibration-of-uncertainty-quantification-
- learning-what-to-evaluate-correlation-aware-decoupling-for-m
- online-sharp-calibrated-bayesian-optimization
- regret-based-epsilonelta-optimal-stopping-criteria-for-bayes
Stable Audio 3
Mentioned in 5 entries
- announcing-the-release-of-stable-audio-3
- running-those-live-lofisynthwave-channels-on-youtube-has-bec
- stable-audio-3
- synchrotron-audioreactive-text2video-stable-audio-3-ltx-23
- system-override-stable-audio-3-ltx-23
Stable Audio 3.0
Mentioned in 5 entries
- local-ai-news-you-missed---may-2026
- meet-stable-audio-30-the-model-family-built-for-artistic-exp
- sharing-a-new-way-to-work-with-stable-audio
- stability-ai-releases-a-new-family-of-audio-models-called-st
- stable-audio-30-is-now-day-0-supported-in-comfyui-open-weigh
Llama 70B
Mentioned in 5 entries
- a-multi-framework-comparison-of-outline-stages-in-long-form-
- budget-aware-compression-pipeline-for-single-gpu-llm-inferen
- dual-3090s
- how-do-you-plan-to-run-qwen38-24t-a95b-locally
- what-options-exist-for-running-the-largest-local-models-at-f
Llama 8B
Mentioned in 5 entries
- a-llamacpp-pr-makes-q20-3036x-faster-on-x86-cpus-8b-decode-g
- dual-3090s
- local-llms-for-use-with-obsidian-vault
- taalas-buried-the-lede-for-the-amazing-demo-of-their-first-t
- your-model-fits-in-vram-your-context-doesnt-the-kv-cache-mat
Flux Klein 9b
Mentioned in 5 entries
- flux-klein-9b-comic-character-lora
- flux-klein-t2i-standalone-app-9b-4b---basic-ai-installations
- gta-70s---teaser-trailer-alternative-version-z-image-turbo--
- gta-70s---teaser-trailer-z-image-turbo---flux-klein-9b---wan
- ltx-23-omninft-flux-klein-9b-via-my-pallaidium-add-on-for-bl
CycleGAN
Mentioned in 5 entries
- application-research-of-a-deep-learning-model-integrating-cy
- counterfactual-explainability-framework-with-cyclegan-and-co
- cranio-diff-diffusion-based-cross-domain-craniofacial-recons
- dsa-cyclegan-a-domain-shift-aware-cyclegan-for-robust-multi-
- unholy-abomination-cyclegan
Ollama
Mentioned in 5 entries
- claude-code-with-ollama-model-and-telegram-channel
- claude-failed-mid-session-ollama-continued-and-handled-the-r
- how-to-add-specific-knowledge-to-an-ollama-model
- released-v080-of-my-local-first-ai-agent-automation-platform
- trooper-load-testing-claude-ollama-tem-behavior-is-consisten
SVM
Mentioned in 5 entries
- benchmarking-logistic-regression-svm-and-lightgbm-against-bi
- benchmarking-logistic-regression-svm-naive-bayes-and-indober
- ibis-a-hybrid-inception-bilstm-and-svm-ensemble-for-robust-d
- multimodal-cultural-heritage-architectural-style-classificat
- toward-near-real-time-marine-oil-spill-detection-in-sar-imag
BitNet
Mentioned in 5 entries
- 100b-models-on-cheap-hardware-how-realistic-and-limitations
- is-ternary-158-bit-llms-making-a-come-back
- spent-two-weeks-on-a-kernel-that-benchmarked-29x-faster-end-
- termezo-ternary-sparse-zeroth-order-optimization-for-fine-tu
- we-applied-bitnet-style-ternary-quantization-to-a-super-reso
LLaDA
Mentioned in 5 entries
- dsl-llada-scaling-continuous-denoising-to-8b-masked-diffusio
- machine-unlearning-for-masked-diffusion-language-models
- pace-dllm-elastic-block-decoding-via-confidence-cliff-estima
- prompt-compression-in-diffusion-large-language-models-evalua
- targeted-remasking-replacing-token-editing-with-token-to-mas
LeJEPA
Mentioned in 5 entries
- dale-ct-depth-aware-foundation-models-for-computed-tomograph
- identifiability-is-what-it-means-to-learn-the-right-latent-s
- object-centric-lejepa
- what-does-jepa-actually-learn-we-can-finally-prove-it-so-exc
- when-does-lejepa-learn-a-world-model
Mythos-class
Mentioned in 5 entries
- anthropic-says-it-expects-mythos-class-models-to-be-availabl
- anthropic-says-red-team-tests-of-fable-5-found-no-universal-
- fable-5-is-something-to-pay-attention-to-this-is-another-i-h
- new-anthropic-introduces-claude-fable-5-a-mythos-class-model
- your-organization-is-not-spending-enough-of-its-efforts-on-b
Kimi-K3
Mentioned in 5 entries
- airllm---recent-updates---with-qwen38-27b-kimi-k3-too
- all-oneshots-from-kimi-k3-looks-better-than-opus48
- b10448
- b10853
- quantized-kimi-k3cloud
Vision-language models
Mentioned in 5 entries
- duomind-enabling-distributed-multi-robot-coordination-with-s
- object-hallucination-free-reinforcement-unlearning-for-visio
- report-supervision
- sole-r1-video-language-reasoning-as-the-sole-reward-for-on-r
- test-time-logit-prompting-for-source-free-missing-modality-a
DDIM
Mentioned in 5 entries
- improved-ddim-sampling-with-moment-matching-gaussian-mixture
- inverting-the-generation-process-of-denoising-diffusion-impl
- one-inverse-step-is-a-convex-program-bayes-limit-calibration
- spatially-adaptive-noise-injection
- why-ddim-hallucinates-more-than-ddpm-a-theoretical-analysis-
Diffusion Policy
Mentioned in 5 entries
- block-wise-adaptive-caching-for-accelerating-diffusion-polic
- dissecting-motion-prior-regularization-for-data-scarce-robot
- how-vlas-fail-differently-black-box-action-monitoring-reveal
- skillmemo-expert-guided-skill-memory-framework-for-compositi
- sparse-actiongen-accelerating-diffusion-policy-with-real-tim
Shieldstral
Mentioned in 5 entries
- introducing-shieldstral-mistrals-3b-open-weights-model-for-c
- mistral-introduces-shieldstral-to-provide-lightweight-policy
- mistral-releases-shieldstral-a-3b-multimodal-safety-classifi
- shieldstral
- the-model-takes-moderation-policy-as-a-plain-language-questi
DreamerV3
Mentioned in 5 entries
- civa-critic-induced-value-subspace-attacks-on-visual-world-m
- correcting-a-learned-physical-invariant-improves-world-model
- dreaming-smoothly-and-sample-efficiently-with-gradient-penal
- reading-the-room-implicit-confusion-encoding-in-recurrent-wo
- the-world-model-remembers-the-actor-forgets-dream-rehearsal-
VAR
Mentioned in 5 entries
- optimizing-2d-input-representations-and-sub-phase-fusion-str
- stylevar-controllable-image-style-transfer-via-visual-autore
- synvar-synergizing-spatial-and-semantic-alignment-in-visual-
- varestorer-one-step-var-distillation-for-real-world-image-su
- where-to-refine-when-to-stop-rethinking-redundancy-via-laten
Moshi
Mentioned in 5 entries
- causal-analysis-and-mitigation-of-spurious-onsets-in-full-du
- environmental-footprint-of-genai-research-insights-from-the-
- integrating-facial-generation-into-full-duplex-spoken-dialog
- metronome-bound-the-cache-keep-the-beat-for-real-time-intera
- using-the-mimi-codec-for-metalinguistic-representations
ConvLSTM
Mentioned in 5 entries
- a-sequence-to-sequence-convlstm-approach-for-leaf-area-index
- benchmarking-convlstm-for-one-day-ahead-imdaa-rainfall-field
- field-validation-of-a-multi-resolution-convlstm-framework-fo
- self-supervised-convlstm-for-fermi-large-area-telescope-tran
- spatio-temporal-forecasting-of-retaining-wall-deformation-mi
GPT-5.2
Mentioned in 5 entries
- best-visual-reasoning-model-in-2026-including-apis-d
- it-is-going-to-be-like-what-happened-in-coding-as-soon-as-mo
- seems-gpt-52-reaches-expert-level-in-peer-review-45-scientis
- study-shows-ai-chooses-nuclear-war-in-crisis
- this-paper-by-researchers-from-mit-and-stanford-finds-that-m
Vision Language Models
Mentioned in 5 entries
- embodied3dbench-benchmarking-low-level-embodied-spatial-inte
- moving-beyond-diversity-visual-token-pruning-as-subspace-rec
- see-the-past-time-reversed-scene-reconstruction-from-thermal
- spacetools-tool-augmented-spatial-reasoning-via-double-inter
- when-simpler-is-better-evaluating-translation-pipelines-for-
Kimi K2.7
Mentioned in 5 entries
- cognition-releases-swe-17-trained-from-kimi-k27-and-availabl
- for-coding-agents-trustworthiness-has-to-be-tested-in-the-ha
- introducing-ai-browser-games-watch-open-closed-models-build-
- try-kimi-k27-and-glm-52-for-free-in-devin-desktop-and-cli
- yesterday-we-launched-swe-17-built-on-the-open-source-kimi-k
Qwen 3.8-Max
Mentioned in 5 entries
- a-quick-snapshot-of-where-qwen38-max-stands-today-qwen38-max
- ask-anything-anonymously-qwen38-max-has-landed-on-venice-giv
- qwen-38-max-is-live-on-modal-with-the-full-1m-context-window
- qwen-38-max-the-latest-model-from-alibabaqwen-is-now-availab
- the-cloud-was-always-a-cat-qwen38-max-just-saw-it-first-try-
Muse Spark 1.3
Mentioned in 5 entries
- artificial-analysis-coding-agent-index-gpt-6-astra-scored-67
- cant-wait-for-the-muse-spark-open-weight-releases-thank-you-
- meta-says-it-has-caught-up-with-anthropic-and-openai-with-mu
- meta-the-frontier-ai-lab
- metas-personal-ai-agent-muse-is-powered-by-muse-spark-13-and
Sonnet 4.6
Mentioned in 5 entries
- 5x-faster-than-llms-while-being-13-cheaper-than-gpt-56-luna-
- big-theme-of-2026---cost-of-closed-models-is-too-high-really
- deepseek-v4-flash-0731-ranks-21-on-agent-arena
- mai-1-has-no-independent-benchmarks-yet-but-the-ones-they-re
- tis-the-year-of-open-source-llms-in-agents
Qwen3.5-27B
Mentioned in 5 entries
- bonsais-document-reveal-how-much-cherry-picked-their-headlin
- lm-performanceqwen36-35b-a3b-outperforms-the-dense-27b-param
- the-llm-tunes-its-own-llamacpp-flags-54-toks-on-qwen35-27b
- thinkingcap-qwen36-27b-warrants-a-look
- tudum-a-turkish-thinking-reasoning-pipeline-for-qwen35-27b
Rampart
Mentioned in 5 entries
- full-interview-engineers-edward-coristine-as400495-and-tai-g
- rampart-from-the-ndstudio-whitehouse-is-number-one-trending-
- rampart-our-pii-removal-model-has-cracked-the-first-screen-o
- tired-the-us-government-regulating-open-source-ai-models-wir
- today-we-are-releasing-rampart-a-147mb-machine-learning-mode
Grok Build
Mentioned in 5 entries
- grok-build-composer-25-so-i-had-an-issue-with-a-swift-app-on
- holy-shit-im-going-to-need-another-grok-heavy-account-grok-b
- i-like-how-grok-build-steadily-updates-the-experiment-notes-
- just-spent-a-full-coding-session-with-grok-build-and-honestl
- we-will-continue-to-make-refinements-to-the-grok-build-harne
GPT 5.6 Luna
Mentioned in 5 entries
- 5x-faster-than-llms-while-being-13-cheaper-than-gpt-56-luna-
- ainews-openai-launches-gpt-56-solterraluna-codex-becomes-cha
- congrats-to-openai-on-the-highly-anticipated-launch-of-gpt-5
- gpt-56-sol-luna-and-terra-now-available-on-ai-gateway
- i-benchmarked-qwen38-27b-on-browser-tasks-its-on-par-with-gp
Qwen 3.6-27B
Mentioned in 5 entries
- deepseek-v4-flash-vs-qwen36-27b-35-122b-and-gemma-4-31b-benc
- qwen-image-bench---finetune-for-image-eval
- tensorrt-edge-llm-completes-the-mlperf-edge-agentic-benchmar
- the-new-qwen36-27b-just-gave-me-definitely-the-best-pelican-
- why-there-is-no-cloud-version-for-qwen-36-2735b
Opus 4
Mentioned in 5 entries
- anthropic-details-how-it-improved-claudes-safety-training-af
- as-shocking-as-the-kimi-k3-release-massive-performance-gain-
- grok-45-context-window-will-upgrade-to-1m-probably-by-next-w
- make-sure-to-read-the-blog-post-for-a-detailed-analysis-of-f
- physical-foundation-models-fixed-hardware-implementations-of
Gemini Pro
Mentioned in 5 entries
- 1-on-memory-benchmark-longmemeval-with-gemini-flash-not-pro-
- calling-all-builders-google-ai-pro-and-ultra-subscribers-wil
- map-of-europe-by-geminis-pro-model
- the-gemini-pro-models-do-not-seem-to-be-iterating-anywhere-n
- the-team-tryheidi-didnt-want-to-keep-renting-someone-elses-i
OpenCode
Mentioned in 5 entries
- harness-showdown-claude-code-vs-opencode-vs-pi-with-deepseek
- how-do-you-deal-with-long-context-sessions-after-restarting-
- local-llms-for-ai-agents
- nexus-connects-to-the-agentic-harnesses-your-teams-already-u
- we-are-open-sourcing-litellm-agent-platform-a-self-hosted-os
Gemini Live
Mentioned in 5 entries
- did-you-know-that-you-can-use-the-googleaistudio-gemini-live
- if-you-are-building-real-time-voice-agents-with-googledeepmi
- read-the-blog-to-learn-more-httpsbloggoogleinnovation-and-ai
- so-cute
- weve-partnered-with-huggingface-and-pollenrobotics-to-bring-
DeepSeek V4 Pro 0813
Mentioned in 5 entries
- deepseek-v4-pro-0813-is-now-live-in-bionic-this-model-shows-
- deepseek-v4-pro-0813-on-openrouter
- deepseek-v4-pro-0813-vs-claude-fable-5-on-deepswe-cost-codin
- deepseek-v4-pro-0813-vs-gpt-56-sol-on-deepswe-cost-coding-an
- qwen-38-27b-scores-52-on-the-artificial-analysis-intelligenc
Gemini 3.5 Transcribe
Mentioned in 5 entries
- google-announces-gemini-35-transcribe-for-ai-powered-speech-
- googles-new-ai-transcription-edits-out-your-8216ums8217-and-
- intelligent-transcription-with-gemini-35-transcribe
- microsoft-ai-debuts-mai-transcribe-2-a-speech-recognition-mo
- today-were-introducing-gemini-35-transcribe-our-latest-trans
LFM2.5-2.6B
Mentioned in 5 entries
- a-26b-model-with-tool-calling-and-128k-context-now-runs-at-3
- hey-everyone-i-just-released-a-sub-6b-sparse-activation-ai-m
- lfm25-26b-is-out
- lfm25-26b-modelkv-cache-quantization-report
- lfm25-26b-on-a-oneplus-13-at-17-toks-pure-cpu
GPT-OSS
Mentioned in 5 entries
- command-a-sets-a-new-high-for-coheres-machine-translation-ca
- gpt-oss-has-turned-one-year-old-today
- improved-and-fixed-template-for-gpt-oss-again-includes-prese
- moe-cpu-offload-benchmark-on-deepseek-v4gemma4qwengpt-oss-te
- openai-says-its-jalapeo-chip-delivered-15x-19x-more-ai-work-
NVIDIA Nemotron
Mentioned in 5 entries
- building-an-analysis-ai-agent-for-industrial-alarm-managemen
- evaluate-clinical-asr-models-faster-with-agent-skills-and-nv
- open-models-closed-environments-palantir-brings-secure-ai-to
- run-nvidia-nemotron-and-openai-gpt-oss-models-on-amazon-bedr
- trying-to-optimize-llm-inference-latency-ollama-openwebuiany
GPT-5.6 Sol Ultra
Mentioned in 5 entries
- and-heres-an-even-better-version-built-by-gpt-56-sol-ultra-r
- i-had-codex-desktop-and-gpt-56-sol-ultra-take-a-go-at-buildi
- moonlight-mayhem-raccoon-heist-by-codex-gpt-56-sol-ultra
- this-time-it-is-novel-math-proofs-with-a-public-model-most-o
- yesterday-we-made-gpt-56-sol-ultra-generally-available-today
Claude Haiku 4.5
Mentioned in 5 entries
- amazon-bedrock-expands-claude-model-availability-to-in-count
- claude-in-microsoft-foundry-is-now-generally-available-hoste
- implement-vector-prompt-document-classification-using-amazon
- procedural-skill-sft-across-capacity-tiers-a-w-shaped-pre-sf
- this-thread-is-worth-reading-it-is-both-hilarious-and-a-good
Nemotron 3 Super
Mentioned in 5 entries
- jupiter-n-technical-report
- nemotron-3-super-open-efficient-mixture-of-experts-hybrid-ma
- rl-fine-tuning-is-now-live-for-nvidiaai-nemotron-3-on-firewo
- the-top-ai-papers-of-the-week-april-13---19---alphaeval---ai
- weve-gone-even-farther-nemotron-3-super-is-120b-and-pretrain
Qwen-3.8-27B
Mentioned in 5 entries
- does-anyone-have-real-experience-with-ornith-15-9b-for-codin
- i-benchmarked-ifmk2-horizon-7b-on-16gb-vram
- qwen-38-27b-is-good-enough-that-i-stopped-using-api
- qwen-38-27b-nemotron-35-lightning-30b-a3b-ornith-15-35b-a3b-
- strix-halo-8060s-gfx1151-qwen-38-27b-q8-and-q6-ud-v3-up-to-2
GLM-5.3 Flash
Mentioned in 5 entries
- breaking-glm-53-flash-by-zaiorg-debuts-at-6th-overall-on-des
- glm-53-now-beats-gpt-56-sol-and-claude-fable-5-on-agentic-be
- meet-wally-our-inference-stack-for-open-frontier-models-buil
- ollama-pricing-vs-openrouter
- theres-a-real-chance-glm-53-flash-is-outperforming-deepseek-
DeepSeek v4 Pro
Mentioned in 5 entries
- deepseek-v4-pro-costs-most-but-it-costs-less
- deepseek-v4-pro-is-now-on-ollamas-cloud-try-it-with-claude-c
- heres-deepseek-v4-pro-added-to-the-playable-gallery-as-well
- we-are-now-enabling-a-queue-for-deepseek-v4-pro-expect-longe
- what-is-wrong-with-deepseek-v4-pro
Fugu Max
Mentioned in 5 entries
- fugu-max-is-now-live-on-openrouter-our-learned-multi-agent-o
- fugu-max-pushes-the-pareto-frontier-for-price-performance-2-
- introducing-fugu-max-and-fugu-ultra-v2-orchestrating-the-par
- introducing-fugu-max-and-fugu-ultra-v2-the-next-evolution-of
- sakana-ai-aifugu-maxnvidia-nemotron
Qwen3.8 Flash Next
Mentioned in 5 entries
- how-would-you-go-about-using-multiple-models-together-from-a
- qwen38-flash-next-llamacpp-config-tuning
- qwen38-flash-next-now-at-12k-ts-prefill-on-strix-halo
- r9v-update-created-and-adopted-kva-projections-based-on-deep
- r9v-update-now-100-toks-in-tg-on-qwen38-flash-next-iq4xs-on-
Gemini 3.8 Live Extended Thinking
Mentioned in 5 entries
- gemini-live-audio
- googles-new-speech-model-gemini-38-live-supports-real-time-r
- introducing-our-most-advanced-gemini-audio-models-yet-gemini
- its-time-for-our-end-of-week-recap-gemini-38-live-and-38-liv
- power-your-agents-gemini-38-live-with-live-avatar-is-now-gen
MiMo-V2.6
Mentioned in 5 entries
- b11102
- mimo-v26-is-simply-the-best-for-now-despite-its-simple-archi
- mimo-v26-update-tool-call-repetition-diagnosed-fixed-after-t
- v050
- xiaomi-introduces-mimo-v26-series-open-source-ai-model-famil
Sonnet 5.5
Mentioned in 5 entries
- anthropic-releases-sonnet-55-saying-it-generates-outputs-30-
- claude-codes-next-era-thariq-shihipar-anthropic
- sonnet-55-is-by-far-the-best-free-model-available-right-now
- the-most-important-thing-about-sonnet-55-is-that-its-now-the
- while-the-token-prices-for-sonnet-55-remain-the-same-as-sonn
ChatGPT Enterprise
Mentioned in 4 entries
- bbva-chatgpt-enterprise-2025-rollout
- bbva-chatgpt-enterprise-bank-rollout
- cyberagent-moves-faster-with-chatgpt-enterprise-and-codex
- openai-helps-hyatt-advance-ai-among-colleagues
Llama 3.1 8B
Mentioned in 4 entries
- benchmarked-qwen-3-8b-llama-31-8b-qwen-25-7b-vs-gpt-4o-on-m4
- llama-31-8b-nails-sdq-but-completely-chokes-on-mdq-20k-token
- llama31-8b-performance-on-a-i71070ti-8gb
- recommended-model-for-a-4060ti-8gb-and-16gb-ram
Kimi-K2.5
Mentioned in 4 entries
- best-visual-reasoning-model-in-2026-including-apis-d
- dflash-for-kimi-k25-was-pushed-3-hours-ago-acceptance-length
- this-is-big-local-ai-news-a-new-open-source-computer-use-llm
- v0205-rc1
VibeVoice
Mentioned in 4 entries
- a-fully-character-driven-fantasy-story-made-entirely-with-lt
- kugel-2-model-vibevoice-finetune-repo-is-gone-does-anyone-kn
- microsofts-mit-licensed-vibevoice-speech-to-text-model-think
- microsoftvibevoice
Qwen Image Edit
Mentioned in 4 entries
- advanced-inpaintedit-kleinqwen-workflows
- qwen-edit-vs-flux-klein
- qwen-image-edit-always-adds-visible-or-protruding-ribs-to-ev
- why-does-qwen-image-edit-shit-itself-when-trying-to-generate
DeepSeek-V3
Mentioned in 4 entries
- deepseek-just-solved-ais-billion-dollar-problem
- gqla-group-query-latent-attention-for-hardware-adaptive-larg
- strong-release-glm-51-is-a-deepseek-v32-like-architecture-in
- the-new-deepseek-v4-like-deepseek-v3-uses-concepts-from-our-
Claude Sonnet 4.5
Mentioned in 4 entries
- emotion-concepts-and-their-function-in-a-large-language-mode
- replicating-the-geometry-of-emotion-representations-in-a-bas
- so-right-now-basically-anyone-with-a-16gb-vram-card-can-go-o
- where-do-models-find-happiness-emotion-vectors-in-open-sourc
Qwen Code
Mentioned in 4 entries
- comparing-how-cline-kilo-and-qwen-code-handle-long-task-cont
- cron-jobs-ai-that-works-on-a-schedule-tell-qwen-code-check-i
- lightweight-agent-to-use-on-laptop
- local-llms-for-ai-agents
DNN
Mentioned in 4 entries
- defending-against-patch-based-and-texture-based-adversarial-
- ensemble-deep-learning-models-for-early-detection-of-meningi
- evolutionary-hyperparameter-optimization-to-find-lightweight
- predictions-of-charge-density-distributions-for-nuclei-with-
Ernie Turbo
Mentioned in 4 entries
- comparing-realism-z-image-turbo-vs-ernie-turbo-vs-klein-9b--
- ernie-turbo-images---res2m-bongtangent-image-to-image-at-41-
- round-3-of-me-fighting-the-grid-on-ernie-turbo
- trying-to-accomplish-realism-with-ernie-turbo---heres-what-i
Neural Processes
Mentioned in 4 entries
- npboost-neural-processes-with-gradient-boosted-fixed-effects
- on-the-conditioning-consistency-gap-in-conditional-neural-pr
- scalable-spatiotemporal-inference-with-biased-scan-attention
- three-costs-of-amortizing-gaussian-process-inference-with-ne
FSD v14.3.1
Mentioned in 4 entries
- fsd-v1431-did-an-exceptional-job-with-parking-at-costco-now-
- fsd-v1431-drives-me-45-minutes-through-la-rush-hour-with-zer
- fsd-v1431-review-after-10-drives-and-many-hours-here-are-my-
- wow-fsd-v1431-drove-me-to-5-destinations-that-grok-inputted-
SIREN
Mentioned in 4 entries
- fire-frequency-reparameterization-as-a-preconditioner-for-pe
- fm-siren-fm-finer-implicit-neural-representation-using-nyqui
- generating-bad-apple-autonomously-from-a-single-initial-stat
- how-much-of-the-weight-space-perception-gap-is-actually-symm
SPLADE
Mentioned in 4 entries
- from-tokens-to-concepts-leveraging-sae-for-splade
- sparse-index-posting-data-bm25-exact-matches-splade-transfor
- the-pre-training-study-of-expanded-splade-models-on-web-docu
- the-role-of-vocabularies-in-learning-sparse-representations-
Grok Voice
Mentioned in 4 entries
- grok-voice-is-1
- grok-voice-is-used-by-starlink
- grok-voice-is-used-by-starlink-right-now
- grok-voice-ranked-first
Qwen 3.6-35B
Mentioned in 4 entries
- a-quick-coding-capability-test4-qwen-36-35b-gguf-variants
- help-needed-to-use-ollama-qwen36-35b-a3b-q4km-as-the-model-f
- new-on-ollama-batiaiqwen36-35b-full-qwen-36-lineup-with-tool
- why-there-is-no-cloud-version-for-qwen-36-2735b
FLUX.2 klein 9B
Mentioned in 4 entries
- flux-klein-9b-in-easy-diffusion
- flux-klein-9b-upscaler
- struggling-choosing-one-edit-model-from-klein-9b-or-qwen-251
- try-this-prompt-in-flux-2-klein-9b-ernie-image-turbo-and-z-i
Cosmos
Mentioned in 4 entries
- anima-vs-illustrious
- grok-imagines-video-agent-moment-cosmos-xai-world-models-gen
- lower-the-cost-of-building-and-running-visual-ai-agents-with
- with-nemotron-cosmos-nvida-gonna-commoditise-everyones-compl
Chroma1-HD
Mentioned in 4 entries
- character-workflow-chroma1-hd-flux2-dev-wan-22-ltx-23
- chroma1-hd-character-transfer-with-flux2-dev
- chroma1-v41-v48-radiance-delivering-a-look-similar-to-midjou
- good-training-settings-for-chroma1-hd
Z-Image Base
Mentioned in 4 entries
- famegrid-checkpoint-zib
- realistic-selfie-prompts-for-z-image-turbobase
- same-prompt-on-open-source-models-z-image-base-distilled-kle
- two-staged-workflow-zib-to-zit
GPT Image 2.0
Mentioned in 4 entries
- access-gpt-image-20-natively-in-hermes-agent-update-now-to-g
- gpt-image-20-just-dropped-in-comfyui-via-partner-nodes-this-
- inner-monologue-testing-out-some-new-transition-ideas-made-i
- some-realism-attempts-with-gpt-image-20
Z-Image-Turbo
Mentioned in 4 entries
- how-to-generate-the-exact-same-scene-across-multiple-images-
- super-detailed-comparaison-between-klein-4b-nucleus-image-z-
- your-favourite-z-image-turbo-checkpoints-and-loras
- z-image-fun-lora-distill-with-z-image-turbo
LVLM
Mentioned in 4 entries
- cfpo-counterfactual-policy-optimization-for-multimodal-reaso
- rethinking-reinforcement-fine-tuning-in-lvlm-convergence-rew
- tarac-mitigating-hallucination-in-lvlms-via-temporal-attenti
- ttsd-far-test-time-self-distillation-with-fisher-anchored-re
SenseNova-U1
Mentioned in 4 entries
- sensenova-u1-just-dropped-native-multimodal-genunderstanding
- sensenova-u1-portrait-test---quality-is-not-great-for-photor
- sensenova-u1-technical-report-vae-free-pixel-level-flow-matc
- sensenova-u1-unifying-multimodal-understanding-and-generatio
Qwen 3.5 4B
Mentioned in 4 entries
- best-model-3b-for-multilingual-understanding-instruction-fol
- i-just-made-the-most-expressive-real-time-motion-harness-for
- ollama-qwen3-54b-troubleshooting
- qwen-35-4b-iq2xs-1667-reasoning-performance-from-tensor-leve
Logistic Regression
Mentioned in 4 entries
- benchmarking-logistic-regression-svm-and-lightgbm-against-bi
- benchmarking-logistic-regression-svm-naive-bayes-and-indober
- hybrid-tf--idf-logistic-regression-and-mlp-neural-baseline-f
- securescan-an-ai-driven-multi-layer-framework-for-malware-an
HunyuanOCR
Mentioned in 4 entries
FNO
Mentioned in 4 entries
- advanced-flood-prediction-with-physics-guided-deep-learning-
- bridging-spectral-operator-learning-and-u-net-hierarchies-sp
- flume-fno-data-efficient-and-scalable-prediction-of-3d-wind-
- helix-fno-spectral-domain-operator-learning-coupled-with-a-h
Ideogram4
Mentioned in 4 entries
- crt-screen-game-on-ideogram4-on-civitai
- got-bored-and-vibe-coded-an-improved-lora-loader-ideogram4-f
- ideogram4-how-to-get-photographs
- testing-ideogram4-lora-training
Large Language Model
Mentioned in 4 entries
- agentic-ai-driven-immersive-simulation-a-knowledge-aware-vir
- hpro-hierarchical-progressive-reward-optimization-via-prefer
- semantic-lenia-emergence-of-homeostatic-solitons-within-the-
- visual-instruction-tuning-aligns-modalities-through-abstract
LLAMA
Mentioned in 4 entries
Recurrent Neural Networks
Mentioned in 4 entries
- exploitation-of-hidden-context-in-dynamic-movement-forecasti
- impact-detection-in-fall-events-leveraging-spatio-temporal-g
- non-linear-mechanical-field-reconstruction-coupling-recurren
- recurrent-reinforcement-learning-with-memoroids
Ornith
Mentioned in 4 entries
- 35b-a3b-tool-calling-benchmark-original-qwen-vs-kat-coder-or
- model-page-httpsollamacomlibraryornith
- run-ornith-with-ollama-ollama-run-ornith-for-coding-use-it-w
- searching-for-38-35b-qwen36-35b-a3b-testing-5-finetunes-vs-b
Flux2
Mentioned in 4 entries
- getting-back-into-ai-art-with-flux2
- ltx-23-is-now-supported-in-comfyui-mesh-for-splitting-models
- sweet-spot-where-flux2-doesnt-get-stuck-during-inpainting-or
- why-do-people-like-flux2-klein-edit-so-much
gpt-oss:20b
Mentioned in 4 entries
- codexs-computer-use-not-working-with-ollama
- found-that-gpt-oss20b-can-hallicunate-extremely-bad
- gpt-oss20b-throughput-regression-after-ollama-upgrading-024-
- running-gpt-oss20b-locally-and-grading-it-head-to-head-again
Autoencoder
Mentioned in 4 entries
- arrythml-an-autoencoder-based-tinyml-approach-for-on-device-
- implementation-of-quantum-implicit-neural-representation-in-
- rapid-finfet-modelling-using-an-autoencoder
- revisiting-energy-based-tabular-anomaly-detection-energy-and
Grok Imagine 1.5
Mentioned in 4 entries
- grok-imagine-15-at-rank-1
- grok-imagine-15-httpslinkscomfyorg4toia0t
- iliad-troy-trailer-made-by-grok-imagine-15-which-was-just-re
- same-prompt-3-models-6-camera-moves-we-put-seedance-20-kling
Qwen2.5
Mentioned in 4 entries
- are-verifier-errors-independent-within-a-grpo-group-evidence
- codefinetuner-fine-tune-a-local-code-autocomplete-model-on-y
- system-report-for-ccl25-eval-task-5-new-dataset-and-lora-fin
- transplanting-inverting-and-preventing-a-misalignment-person
Bonsai 27B
Mentioned in 4 entries
- bonsai-27b-1-bit-dense-llm-running-locally-in-your-browser-u
- bonsai-27b-is-a-big-step-for-local-ai-27b-class-multimodal-c
- huge-if-true-we-are-talking-about-a-27b-multimodal-model-tha
- today-were-announcing-bonsai-27b-the-first-27b-class-model-t
GLM-4.5-Air
Mentioned in 4 entries
- b10603
- i-run-glm-45-air-110b-on-16gb-ram-consumer-machine-and-qwen3
- v030
- you-can-now-use-mtp-in-glm-air
Convolutional Neural Network
Mentioned in 4 entries
- developing-a-foundation-model-for-high-resolution-remote-sen
- protect-the-brain-when-treating-the-heart-a-convolutional-ne
- real-time-detection-of-charge-jumps-in-superconducting-qubit
- testing-general-relativity-through-gravitational-wave-classi
GRU
Mentioned in 4 entries
- inductive-power-grid-cascading-failure-analysis-with-gru-gat
- lk-jam-system-architecture-and-implementation-of-a-real-time
- optimizing-2d-input-representations-and-sub-phase-fusion-str
- sensitivity-analysis-of-gru-lstm-and-transformer-encoder-in-
Qwen4
Mentioned in 4 entries
- api-is-live-on-qwencloud-httpswwwqwencloudcommodelsqwen38-fl
- meet-qwen38-flash-a-multimodal-moe-and-an-early-preview-of-t
- qwen38-flash-next
- qwen38-flash-next-time-to-update-those-benchmarks
Qwen3.7-Plus
Mentioned in 4 entries
- alibaba-releases-qwen37-plus-a-multimodal-proprietary-model-
- alibaba-releases-weights-for-qwen38-models-under-apache-20-l
- introducing-qwen37-plus-a-multimodal-agent-model-that-unifie
- we-promised-open-weights-for-qwen38-now-time-to-meet-them-qw
Qwen3.5-9B
Mentioned in 4 entries
- alert-apple-just-dropped-a-new-model-on-hugging-face-its-a-q
- nanbeige42-3b-im-not-impressed
- qwen35-9b-triple-loop
- taking-qwen35-9b-quants-to-sota-new-lineup-incoming
Muse-Glimmer-30B
Mentioned in 4 entries
- 1-day-in-and-i-feel-okay-saying-muse-glimmer-30b-finally-bea
- new-muse-glimmer-30b-sota-quants---hopefully-a-new-lineup
- qwen-38-27b-nemotron-35-lightning-30b-a3b-ornith-15-35b-a3b-
- taking-qwen35-9b-quants-to-sota-new-lineup-incoming
NemoClaw
Mentioned in 4 entries
- another-big-langchain-week
- industrial-software-leaders-build-secure-autonomous-ai-engin
- langchain-launches-this-week-all-about-open-source-models-an
- nvidia-nemoclaw-flaw-let-attackers-poison-the-model-behind-a
H3
Mentioned in 4 entries
- fastvideo-fasth3-v1-open-source-4-step-sparse-distilled-h3-c
- fizgig-52---combining-two-minimax-training-methods-beats-eit
- minimax-are-issuing-takedowns-on-decensorexplicit-h3-loras
- vdn-h3-on-m5-pro-24gb-up-to-26-faster-h3-generation
Gaussian Process
Mentioned in 4 entries
- bayesian-optimisation-using-product-of-experts-gaussian-proc
- online-adaptive-kernel-mixing-for-gaussian-process-decision-
- simplex-to-euclidean-bijection-for-conjugate-and-calibrated-
- three-costs-of-amortizing-gaussian-process-inference-with-ne
Graph Neural Network
Mentioned in 4 entries
- forecasting-individual-netflows-using-a-predictive-masked-gr
- genie-watermarking-graph-neural-networks-for-link-prediction
- hybrid-lstm-graph-neural-framework-for-robust-financial-frau
- physics-informed-graph-neural-network-surrogates-for-turbule
Gated DeltaNet
Mentioned in 4 entries
- gated-deltanet-has-been-one-of-my-favorite-hybrid-attention-
- leapquant-efficient-linear-attention-with-accurate-recurrent
- treewy-speculative-verification-for-gated-deltanet-hybrids
- why-gated-deltanet-survives-4-bit-quantization-nvfp4-w4a4-fo
Amazon Nova Act
Mentioned in 4 entries
- accelerating-software-delivery-with-agentic-qa-automation-us
- first-orion-accelerates-qa-automation-using-amazon-nova-act
- implementing-synthetic-monitoring-using-amazon-nova-act
- scaling-ux-testing-with-amazon-nova-act-a-new-approach-to-us
Namazu
Mentioned in 4 entries
- advanced-reasoning-and-problem-solving-plus-strong-performan
- sakana-ai-sakana-chat-namazu-namazu
- sakana-chat-just-got-a-big-upgrade-no-login-required-free-to
- we-just-pushed-a-big-update-to-sakana-chat-no-login-required
GPT-Image 2
Mentioned in 4 entries
- canvas-updates-include---improved-canvas-interactions-more-p
- gpt-image-2-white-smoke-attack
- have-you-tried-the-new-replit-canvas---create-beautiful-ui-d
- image-generation-with-gpt-image-2-videos-with-seedance-and-a
MARBERT
Mentioned in 4 entries
- arassm-a-bidirectional-state-space-encoder-for-arabic-masked
- does-machine-know-interpersonal-pragmatics-evidence-from-mar
- machine-learning-and-digital-pragmatics-which-word-category-
- spam-and-sentiment-detection-in-arabic-tweets-using-marbert-
Amazon Nova 2 Lite
Mentioned in 4 entries
- how-uniopen-customized-amazon-nova-to-their-retail-moderatio
- object-detection-with-amazon-nova-2-lite
- pair-nova-2-lite-with-claude-for-cost-optimized-document-pro
- prompting-amazon-nova-2-for-content-moderation
Chronos-2
Mentioned in 4 entries
- how-decathlon-runs-demand-forecasting-at-scale-with-chronos-
- investigating-simple-target-covariate-relationships-for-chro
- knowledge-graph-augmented-chronos-2-for-hec-ras-surrogate-fo
- pooling-helps-learned-weighting-hurts-in-context-decomposing
PiD
Mentioned in 4 entries
- nvidia-pid-based-img-upscaler-no-workflow-but-py
- nvidia-pid-preview-inside-a-next-gen-tiled-upscaler-enhancer
- nvidia-pid-upscale-for-video
- pit-nvidia-vs-seedvr2
MiniMax-M3
Mentioned in 4 entries
- first-technical-deepdive-on-m3-on-the-internet
- minimax-m3-combines-1m-context-native-multimodality-and-mini
- minimax-m3-will-by-arrive-on-huggingface-openweight-at-next-
- serving-minimax-m3-for-efficient-inference-unlocking-1m-toke
NVIDIA Cosmos 3
Mentioned in 4 entries
- build-a-physical-ai-model-factory-with-nvidia-cosmos-3-on-sa
- develop-physical-ai-reasoning-world-and-action-models-with-n
- post-train-nvidia-cosmos-3-in-one-day-using-agent-skills
- welcome-nvidia-cosmos-3-the-first-open-omni-model-for-physic
Multimodal Diffusion Transformers
Mentioned in 4 entries
- diagnosing-and-correcting-concept-omission-in-multimodal-dif
- fairflow-demystifying-and-mitigating-stereotype-bias-in-text
- prompt-reinjection-alleviating-prompt-forgetting-in-multimod
- semantic-steering-for-controllable-generation-tuning-free-co
Galactica
Mentioned in 4 entries
- i-have-absolutely-never-ever-asked-to-stop-rd-on-llms-in-fac
- remember-october-2022-when-you-doused-galactica-with-vitriol
- room-2016-for-those-attending-aidotengineer-225pm-will-also-
- thats-right
ChatGPT 5.4
Mentioned in 4 entries
- cant-agree-more-after-trying-out-chatgpt-5455-for-two-months
- chatgpt-54-solved-a-64-year-old-math-problem
- claude-opus-47-was-so-bad-i-temporarily-switched-to-chatgpt-
- did-anyone-suddenly-lose-access-to-54-and-55
cGAN
Mentioned in 4 entries
- a-novel-approach-for-the-counting-of-wood-logs-using-cgans-a
- a-paired-point-of-care-ultrasound-dataset-for-image-quality-
- generative-deep-learning-for-computational-destaining-and-re
- property-constrained-3d-porous-media-reconstruction-from-2d-
Grounding DINO
Mentioned in 4 entries
- abra-teleporting-fine-tuned-knowledge-across-domains-for-ope
- reward-guided-semantic-evolution-for-test-time-adaptive-obje
- unlocking-the-potential-of-grounding-dino-in-videos-paramete
- vlod-tta-test-time-adaptation-of-vision-language-object-dete
Grok 5
Mentioned in 4 entries
- by-the-end-of-the-year-we-should-have-gpt-6-fable-55-gemini-
- elon-last-year-grok-5-has-a-10-chance-of-becoming-worlds-fir
- is-xai-dead-or-what-where-ist-grok-5-lol
- quoting-spacex-s-1
Doubao
Mentioned in 4 entries
- bytedances-doubao-and-alibabas-qwen-will-disable-humanlike-a
- chinas-ev-makers-are-competing-on-in-car-ai-features-bytedan
- icebreaker-for-conversational-agents-breaking-the-first-mess
- sources-qualcomm-reached-a-deal-with-bytedance-to-supply-mil
RT-DETR
Mentioned in 4 entries
- benchmarking-resnet-backbones-in-rt-detr-impact-of-depth-and
- depth-aware-pothole-detection-using-yolo-and-rt-detr-at-the-
- pattern-enhanced-rt-detr-for-multi-class-battery-detection
- pillardetr-yolo-backbone-and-rt-detr-head-for-real-time-3d-o
State Space Models
Mentioned in 4 entries
- design-principles-for-sequence-models-via-coefficient-dynami
- following-motion-for-sequential-modeling-in-video-frame-inte
- rotrnn-modelling-long-sequences-with-rotations
- tcp-ssm-efficient-vision-state-space-models-with-token-condi
DeBERTa
Mentioned in 4 entries
- duluth-at-semeval-2026-task-6-deberta-with-llm-augmented-dat
- fine-tuning-over-architectural-complexity-broad-coverage-pii
- nlpcc-2026-task-10-citation-level-faithfulness-verification-
- shrinepeeuca-2026-rakshak---multi-task-deberta-with-rational
Fugu-Ultra
Mentioned in 4 entries
- fugu-ultra-is-now-available-on-vercel-ai-gateway-httpsvercel
- fugu-ultra-is-now-live-on-openrouter-we-share-a-core-vision-
- model-orchestration-is-in-many-ways-the-natural-outgrowth-of
- openrouter-httpsopenrouteraisakanafugu-ultra
Naive Bayes
Mentioned in 4 entries
- a-blended-likelihood-approach-for-achieving-fairness-using-n
- benchmarking-logistic-regression-svm-naive-bayes-and-indober
- llms-or-naive-bayes-old-gems-or-new-ways
- on-generalized-naive-bayes-with-continuous-features
Step 3.5 Flash
Mentioned in 4 entries
- 10-days-wasnt-enough-step-35-flash-is-back-on-nousresearch-p
- step-35-flash-from-stepfunai-is-now-free-again-in-nous-porta
- step-35-flash-is-free-on-nous-portal-for-another-week-enjoy-
- step-35-flash-is-now-live-for-nous-portal-users-free-for-the
AlphaEvolve
Mentioned in 4 entries
- effective-harness-engineering-for-algorithm-discovery-with-c
- improving-the-matrix-multiplication-exponent-with-modern-opt
- solve-harder-problems-with-alphaevolve-now-available-to-ever
- teaching-llms-to-self-evolve-cultivating-core-meta-skills-wi
ReLU
Mentioned in 4 entries
- cryptanalytic-extraction-of-isolated-bias-free-glu-feed-forw
- deep-ice-the-first-globally-optimal-algorithm-for-minimizing
- intelligence-per-picojoule-with-itsclivetime-and-dylan522p-0
- neural-networks-and-virtual-extended-formulations
Spiking Transformers
Mentioned in 4 entries
- closing-the-theory-practice-gap-in-spiking-transformers-via-
- elastic-spiking-transformers-for-efficient-gesture-understan
- rethinking-attention-locality-in-spiking-transformers
- sage-surrogate-gradient-adaptation-via-attention-guided-entr
DistilBERT
Mentioned in 4 entries
- adversarial-robustness-of-phishing-email-detection-a-compara
- an-explainable-distilbert-bilstm-attention-framework-for-bin
- cognitive-linguistic-indicators-of-depression-in-online-comm
- fine-tuned-a-distilbert-with-ml-intern-today-for-the-first-t
GAT
Mentioned in 4 entries
- cross-modal-emotion-understanding-a-transformer-gat-approach
- explaining-spatial-information-flow-in-short-term-traffic-fo
- gatnexthop-a-gat-for-shortest-path-routing-with-cross-topolo
- when-graph-structure-becomes-a-liability-a-critical-re-evalu
Laguna
Mentioned in 4 entries
- great-to-see-poolsideai-us-lab-committing-to-open-sourcing-t
- question-about-quant-versus-size
- software-engineers-do-you-honestly-get-anything-useful-out-o
- v0323
Muse Spark 1.2
Mentioned in 4 entries
- introducing-muse-code-and-muse-spark-12
- meta-is-offering-a-cheaper-muse-spark-12-contributor-tier-pr
- meta-releases-muse-code-in-beta-a-terminal-coding-agent-powe
- open-source-is-so-back-zuck-just-announced-meta-is-opening-t
Nano Banana
Mentioned in 4 entries
- arena-trends-text-to-image-jan-2026-apr-2026-for-most-of-the
- feels-like-google-could-have-been-the-dominating-force-in-ai
- introducing-expanded-coverage-for-image-generation-models-in
- whats-new-in-ai-infrastructure-and-orchestration-this-month
GPT-4.5
Mentioned in 4 entries
- 54-can-actually-be-funny
- 55-in-codex-is-so-good-for-non-coding-tasks-i-keep-assuming-
- 55-xhigh-in-fast-mode-is-really-good-i-think-i-got-psyoped-b
- weakly-general-ai-achieved-at-least-according-to-the-criteri
Muse Image
Mentioned in 4 entries
- as-part-of-metas-muse-image-rollout-instagram-users-with-pub
- meta-launches-image-generation-model-with-coding-search-capa
- meta-says-muse-image-includes-an-invisible-watermarking-syst
- metas-new-muse-image-model-can-pull-other-instagram-users-in
Gemini Robotics-ER 1.6
Mentioned in 4 entries
- boston-dynamics-robot-dog-now-reads-gauges-and-thermometers-
- deepmind-launches-gemini-robotics-er-16-to-meet-precise-phys
- gemini-robotics-er-16-powering-real-world-robotics-tasks-thr
- google-deepmind-introduces-gemini-robotics-er-16-robotic-rea
Claude 4.6
Mentioned in 4 entries
- 46-max-is-pretty-nice-though-good-job-anthropocites
- claude-46-just-counted-letters-correctly
- i-have-found-47-great-for-design-reverted-back-to-46-extende
- in-claude-code-the-default-effort-is-now-xhigh-a-new-level-b
ChatGPT Codex
Mentioned in 4 entries
- anyone-know-if-chatgpt-codex-in-the-new-chatgpt-desktop-app-
- chatgpt-work-claude-cowork-chatgpt-codex-claude-code-i-kinda
- i-am-so-confused-by-chatgpt-v-chatgpt-codex-v-chatgpt-work-v
- using-blender-with-coding-agents-on-macos
GPT-5.6-Sol
Mentioned in 4 entries
- cors-chat
- dominic-kundel-from-openai-gives-the-inside-scoop-on-where-g
- i-do-love-rottweilers
- openai-and-anthropic-have-both-just-posted-about-an-overlapp
Qwen-Image-3.0
Mentioned in 4 entries
- authentic-details-and-deep-knowledge-are-also-major-upgrades
- meet-qwen-image-30-the-third-generation-of-our-foundational-
- qwen-image-30-pro-is-live-on-qwen-cloud-now-try-it-out-https
- try-qwen-image-30-on-openartai
Gemini 3 Pro
Mentioned in 4 entries
- an-agentic-loop-compile-test-profile-revise-helps-gemini-3-p
- gemini-3-pro-was-the-first-model-to-achieve-at-least-23-on-a
- is-standard-paygo-for-gemini-3-pro-unusable-for-scale
- new-paper-worth-reading-gpt-54-nano-plus-a-critic-comparator
Qwen 9B
Mentioned in 4 entries
- a-ultra-lightweight-mini-agent---zero-framework-and-localoll
- if-anyone-is-running-qwen-9b-or-27b-or-35b-and-getting-wrong
- m3-16gb-running-ollama-qwen-9b-is-extremely-slow-10-12-mins-
- same-9b-qwen-weights-191-in-aider-vs-456-with-a-scaffold-ada
DeepSeek V4-Pro
Mentioned in 4 entries
- deepseek-launches-v4-pro-its-most-advanced-model-that-rivals
- deepseek-releases-v41-flash-says-it-outperforms-flagship-v4-
- deepseek-v4-pro-is-live-on-fireworks-frontier-class-coding-a
- deepseek-v4-pro-on-fireworks-zoooooom
Qwen Image
Mentioned in 4 entries
- invokeai-613-just-released-its-largest-community-driven-rele
- invokeai-6130-released
- lora-for-qwen-image-2512
- staging-worflow-qwen
GPT-5.6-Cyber
Mentioned in 4 entries
- openai-gpt-56-cyber-daybreak-expansion
- openai-releases-gpt-56-cyber-a-more-cyber-permissive-version
- were-expanding-our-cybersecurity-initiative-daybreak-and-int
- weve-used-gpt-56-cyber-extensively-in-real-world-vulnerabili
NVIDIA Nemotron 3.5 Lightning
Mentioned in 4 entries
- nvidia-nemotron-35-lightning-delivers-fast-accurate-speciali
- nvidia-nemotron-35-lightning-now-available-in-amazon-sagemak
- portable-computer-runs-a-post-trained-and-locally-installed-
- v0329
Qwen3.8 27b
Mentioned in 4 entries
- mac-heads-is-there-any-point-to-mlx-in-september-2026
- ngl-im-hyped-to-see-if-qwen38-27b-can-make-me-a-sandwich-ins
- qwen38-27b-for-agentic-coding-and-next-what
- whats-the-story-with-agnes-30-flash
Vision-Language Model
Mentioned in 4 entries
- cross-domain-tracker-adaptation-without-target-domain-labels
- decoupling-vision-language-and-action-for-efficient-multi-ta
- efficient-vision-language-action-management-and-serving-for-
- the-mirage-of-calibrated-confidence-trajectory-independence-
DeepSeek-OCR
Mentioned in 4 entries
- adversarial-attacks-on-deep-ocr-systems
- b10285
- decoupling-semantics-from-vision-a-framework-for-faithful-vi
- rtprune-reading-twice-inspired-token-pruning-for-efficient-d
Nemotron Ultra
Mentioned in 4 entries
- proud-to-support-the-open-source-community-thanks-for-the-ne
- ssankar-says-palantir-was-able-to-make-nvidias-nemotron-ultr
- thinky-with-a-1t-param-41b-active-apache-2-model-benchmarks-
- us-open-source-models-are-quickly-gaining-ground-nvidias-new
Gemma 4 E2B
Mentioned in 4 entries
- i-put-gemma-4-e4b-and-e2b-into-an-e-reader-so-i-can-ask-my-w
- mtp-support-for-gemma-4-e2b-and-e4b-assistants-by-max-krasny
- the-model-punches-above-its-weight-outperforming-gemma-4-e2b
- voice-conversations-between-gemma4-12b-and-e2b-on-gpu-and-je
GPT-5.4-Cyber
Mentioned in 4 entries
- openai-launches-gpt-54-cyber-model-for-vetted-security-pros
- openai-rolls-out-gpt-54-cyber-a-fine-tuned-gpt-54-variant-fo
- sources-openai-has-been-briefing-us-federal-agencies-state-g
- were-expanding-trusted-access-for-cyber-with-additional-tier
Qwen Image Edit 2511
Mentioned in 4 entries
- my-steps-and-yours-anima-base-10---qwen-image-edit-2511---wa
- struggling-choosing-one-edit-model-from-klein-9b-or-qwen-251
- the-image-generated-by-qwen-image-edit-2511-is-blurred-anime
- turning-anime-into-real-and-testing-klein9b-vs-qwen-edit-251
Gemini Robotics 2
Mentioned in 4 entries
- for-decades-weve-dreamed-of-robots-that-can-seamlessly-step-
- google-deepmind-debuts-gemini-robotics-2-model-series-for-hu
- googledeepmind-humanoid-legs-or-wheeled-rovers-should-robots
- its-been-a-busy-couple-of-weeks-icymi-heres-the-recap-gemini
Ornith-1.5-35B-A3B
Mentioned in 4 entries
- ornith-15-35b-a3b-on-8-gb-vram-i-think-ive-found-my-sweet-sp
- ornith-15-35b-a3b-q4-running-60tks-on-4070ti
- qwen-38-27b-nemotron-35-lightning-30b-a3b-ornith-15-35b-a3b-
- we-have-q38-35b-at-home-3x-new-ornith-15-released
Gemini 4
Mentioned in 4 entries
- gemini-4-is-almost-ready-says-new-google-deepmind-chief
- google-deepmind-svp-koray-kavukcuoglu-says-gemini-4-is-in-th
- sources-google-plans-to-release-gemini-38-flash-as-soon-as-w
- sources-some-google-employees-say-gemini-4-performs-well-on-
GPT-5.4 Pro
Mentioned in 4 entries
- did-the-100-plan-affect-the-gpt-54-pro-model
- earlier-this-month-an-erds-problem-that-had-been-open-for-60
- openais-gpt-54-pro-reportedly-solves-a-longstanding-open-erd
- this-wasnt-the-case-with-previous-image-generators-but-the-l
Qwen3.5-0.8B
Mentioned in 4 entries
- guarded-gradient-based-activation-steering-of-shutdown-respo
- i-fine-tuned-a-08b-local-model-for-dictation-cleanup-it-matc
- ovisocr2-08b-first-end-to-end-model-to-top-omnidocbench---i-
- qwengram-08b-i-transferred-qwen38-flash-nexts-n-gram-memory-
Nemotron-3-Super
Mentioned in 4 entries
- banger-compression-paper-from-nvidia-bookmark-it-bigger-moe-
- been-having-an-argument-with-nemotron-3-supercloud-today
- model-add-nvidia-nemotron-3-puzzle-75b-a9b-nemotronhpuzzle-s
- nemotron-labs-3-puzzle-75b-a9b-compressing-hybrid-moe-llms
Qwen 3.6 27b
Mentioned in 4 entries
- how-come-artificialanalysisai-ranks-gemma4-above-qwen36-27b-
- qwen-38-35ba3b-wen
- qwen-38-next-flash-is-really-really-really-verbose
- what-are-the-best-models-you-can-run-on-your-nvidiaai-dgx-sp
Veo
Mentioned in 4 entries
- feels-like-google-could-have-been-the-dominating-force-in-ai
- meet-gemini-omni-11-flash-our-newest-multimodal-model-for-vi
- specsem-net-integrating-spectral-and-semantic-features-for-r
- veo-act-enhancing-vla-policies-with-frontier-video-models
NVIDIA Nemotron 3 Ultra
Mentioned in 4 entries
- before-the-week-ends-lets-acknowledge-one-of-the-most-insane
- create-a-langchain-deep-agents-harness-profile-for-nvidia-ne
- introducing-nvidia-nemotron-3-ultra-a-frontier-smart-open-mo
- nvidia-nemotron-3-ultra-is-on-fireworks-day-zero-nemotron-ul
Qwen Flash Next
Mentioned in 4 entries
- build-around-cmp-170hx-for-qwen-flash-next
- qwen-flash-next-mtp-work-restarted
- qwen4exp-add-mtp-by-am17an-pull-request-29761-ggml-orgllamac
- qwen4exp-fixes-in-llamacpp
Gemma-4-31B-it-Pearl
Mentioned in 4 entries
- gemma-4-31b-it-pearl-supports-256k-context-configurable-thin
- gemma-4-31b-it-pearl-supports-32k-context-configurable-think
- gemma-4-31b-it-pearl-supports-text-and-image-input-256k-cont
- introducing-gemma-4-31b-it-pearl-on-together-ai-pearl-resear
Qwen 3.8B
Mentioned in 4 entries
- dear-24g-owners-try-vllm-you-might-be-able-to-run-qwen38-27b
- i-turned-an-asymetric-pair-of-tesla-v100s-pcie-both-16-gb-32
- splash-110-released-gguf-quants-support-mlx-import-and-more
- which-qwen-for-vllm
Qwen 3.5 9B
Mentioned in 4 entries
- how-accurate-do-you-think-this-is-qwen35-9b-vs-gpt-4o
- is-llamacpp-meant-to-be-slow-at-long-context-even-when-you-a
- ollama-qwen-35-9b-cannot-maintain-focus
- psa---comfyui-now-support-system-prompt-for-text-generate-qw
Gemma4 12B
Mentioned in 4 entries
- best-local-models-for-claude-code-with-8gb-vram
- gem16---custom-engine-for-gemma4-12b-26b-on-blackwell-16gb-g
- how-would-you-go-about-using-multiple-models-together-from-a
- voice-conversations-between-gemma4-12b-and-e2b-on-gpu-and-je
Fugu Ultra v2
Mentioned in 4 entries
- fugu-ultra-v2-by-the-numbers-full-results-httpssakanaaifugu-
- introducing-fugu-max-and-fugu-ultra-v2-orchestrating-the-par
- introducing-fugu-max-and-fugu-ultra-v2-the-next-evolution-of
- sakana-ai-aifugu-maxnvidia-nemotron
Ternary Bonsai 2 27B
Mentioned in 4 entries
- bonsais-document-reveal-how-much-cherry-picked-their-headlin
- ternary-bonsai-is-a-headless-chicken
- this-is-basically-an-opus-4546-level-model-that-will-work-on
- today-were-announcing-ternary-bonsai-2-27b-based-on-qwen38-2
DeepSeek v4.1 Flash
Mentioned in 4 entries
- can-you-take-the-cheapest-open-model-from-a-frontier-lab-bea
- deepseek-v41-flash-peak-rates-why
- so-deepseekai-v41-flash-i-estimate-cost-10m-to-train-100x-le
- thank-you-ollama-deepseek-v41-flash-is-the-best
Mica v0.1 4B
Mentioned in 4 entries
- kev-4b-topped-out-in-every-tetris-game-i-ran-mica-v01-4b-cle
- mica-v01-4b-got-an-iron-pickaxe-in-real-minecraft-without-ge
- mica-v01-4b-got-diamonds-in-survival-minecraft-on-its-first-
- mica-v01-4b-open-jev-style-decision-model-yesno-choice-score
MiMo-V2.6-Pro
Mentioned in 4 entries
- mimo-v26-both-pro-and-flash-is-a-benchmaxxed-scam
- mimo-v26-pro-debuts-as-the-top-open-weights-model-on-the-art
- mimo-v26-pro-is-live-on-wally-day-1-xiaomimimos-new-open-wei
- mimo-v26-pro-ties-grok-47-xhigh-and-beats-glm-53-max-on-arti
Gemini 3.8 Flash TTS
Mentioned in 4 entries
- check-out-this-weeks-updates-and-releases-gemini-38-flash-tt
- gemini-38-tts-playground
- google-launches-two-benchmark-topping-speech-generation-mode
- were-launching-gemini-38-flash-tts-and-gemini-38-flash-lite-
Clef
Mentioned in 4 entries
- clef-flash-httpsollamacomlibraryclef-flash-clef-httpsollamac
- cloudflare-debuts-open-weight-multimodal-decision-models-cle
- cloudflares-decision-models-are-available-on-ollama-decision
- introducing-clef-our-open-source-decision-models-and-new-rl-
Mistral 7B
Mentioned in 3 entries
- how-to-fine-tune-an-llm-an-end-to-end-guide
- recommended-model-for-a-4060ti-8gb-and-16gb-ram
- variance-is-not-importance-structural-analysis-of-transforme
SeedVR2
Mentioned in 3 entries
- flux2-klein-2-stage-upscale
- flux2-klein-tiling-upscale-workflow
- implementing-super-resolution-by-deploying-seedvr2-on-amazon
ACE-Step 1.5
Mentioned in 3 entries
- ace-step-15-xl-base-bf16-version-converted-from-fp32
- acestep-xl-tips
- does-ace-step-15-do-lyrics-on-its-own
Amazon Nova Multimodal Embeddings
Mentioned in 3 entries
- building-intelligent-audio-search-with-amazon-nova-embedding
- manufacturing-intelligence-with-amazon-nova-multimodal-embed
- power-video-semantic-search-with-amazon-nova-multimodal-embe
MiMo V2 Pro
Mentioned in 3 entries
- lets-see-how-mimo-v2-pro-evolves-in-hermes-agent
- were-glad-hermes-users-have-been-making-use-of-the-free-mimo
- xiaomi-httpsportalnousresearchcom
Neural Operators
Mentioned in 3 entries
- feature-interaction-modeling-for-physics-informed-neural-net
- meno-meanflow-enhanced-neural-operators-for-dynamical-system
- the-fractal-neural-operator-overcoming-spectral-bias-in-chao
Deep Image Prior
Mentioned in 3 entries
- a-principled-self-referenced-early-stopping-approach-for-dee
- pooling-based-context-modeling-for-convolution-free-deep-ima
- preventing-overfitting-in-deep-image-prior-for-hyperspectral
Waypoint-1.5
Mentioned in 3 entries
- waypoint-15-a-real-time-video-world-model-for-consumer-hardw
- waypoint-15-higher-fidelity-interactive-worlds-for-everyday-
- waypoint-15-new-open-source-world-model-trained-on-fps-games
GLM-5V-Turbo
Mentioned in 3 entries
- glm-5v-turbo-on-ai-gateway
- glm-5v-turbo-tech-report-toward-a-native-foundation-model-fo
- glm-5v-turbo-toward-a-native-foundation-model-for-multimodal
Chat
Mentioned in 3 entries
- having-a-lot-of-fun-with-this-i-gave-chat-a-scenario-you-get
- maybe-i-should-try-chat-again-after-only-using-claude-for-a-
- wow-wonder-how-many-people-on-this-site-hyped-this-paper-and
Large Multimodal Models
Mentioned in 3 entries
- efficient-data-selection-for-multimodal-models-via-increment
- latent-implicit-visual-reasoning
- mitigating-visual-context-degradation-in-large-multimodal-mo
Illustrious XL
Mentioned in 3 entries
- illustrious-noobai-style-explorer-now-with-16000-danbooru-ar
- illustrious-z
- where-to-find-complete-illustriousnoobai-character-keywords
LTX Video 2.3
Mentioned in 3 entries
- acestep-15-xl-turbo-ltx-23-on-an-8gb-rtx-5060-laptop
- can-anyone-tell-me-if-ltx-23-runs-well-on-an-m4-pro-with-24g
- ltx-23-how-to-stop-characters-from-cloning-themselves
Rectified Flow
Mentioned in 3 entries
- a-lagrangian-view-of-flow-matching
- i-have-distilled-my-flow-matching-model-into-the-rectified-f
- mixflow-mixed-source-distributions-improve-rectified-flows
LightX2V
Mentioned in 3 entries
- is-there-a-new-wan22-lightx2v-20260412-fp8-version-for-comfy
- new-wan-22-lightx2v-speed-lora-260412
- where-do-i-download-lightx2vi2v22highsafetensors
DALL·E
Mentioned in 3 entries
- a-school-book-from-the-year-2035-made-by-dalle-are-we-cooked
- hey-does-anyone-know-how-to-make-it-generate-more-detailed-i
- pagedmark-invisible-synthid-class-watermark-removal-for-ai-i
ZIB
Mentioned in 3 entries
Qwen-Image-Edit-2511
Mentioned in 3 entries
- attnrouter-per-category-attention-routing-for-training-free-
- qwen-2511-fp8-mixed-taking-3040s-per-image-edit-which-gguf-s
- vnccs-qie2511-posestudio-lora-for-art-has-been-updated
Flux.2 Klein 4B
Mentioned in 3 entries
- complex-weird-prompt-test-ernie-turbo-flux2-klein-4b-z-image
- flux2-klein-9b-4b-scribbly-doodle-lora
- has-anyone-done-partial-fine-tuning-on-flux2-klein-4b-to-enf
Latent Diffusion Model
Mentioned in 3 entries
- domain-specific-latent-representations-improve-the-fidelity-
- finally-got-around-to-making-a-proper-ldm
- no-prompt-no-leaks-a-robust-generative-steganography-framewo
Nucleus-Image
Mentioned in 3 entries
- nucleus-image-sparse-moe-for-image-generation
- nucleusainucleus-image-samples
- super-detailed-comparaison-between-klein-4b-nucleus-image-z-
Ernie Image
Mentioned in 3 entries
- ernie-image-vs-zimage-base-style-comparison
- i-made-a-patch-for-ernie-image-fp16-support-in-comfyui-for-2
- onetrainer-now-supports-ernie-lora
HY-World 2.0
Mentioned in 3 entries
- comfyui-hy-world2
- comfyuihyworld2-update-quality-improvement-world-stereo-ligh
- were-open-sourcing-hy-world-20-a-multimodal-world-model-that
V-JEPA2
Mentioned in 3 entries
- behavioral-geometric-supervision-aligns-video-foundation-mod
- beyond-the-beep-scalable-collision-anticipation-and-real-tim
- most-physical-ai-models-recognize-patterns-they-dont-underst
Flux.2 Klein
Mentioned in 3 entries
- flux-klein-is-better-than-any-closed-model-for-image-editing
- flux2-klein-prompt-help---cannot-get-rid-of-studio-camera-fl
- flux2klein-little-info
Flux Dev
Mentioned in 3 entries
- flux-dev1-artistic-mix-04-16-2026
- wan-27-vs-flux-dev-on-the-same-character-prompts-clear-winne
- which-model-would-be-the-best-to-generate-fictional-country-
Flux.1
Mentioned in 3 entries
- flux-dev1-artistic-mix-04-16-2026
- local-generation-is-falling-behind
- z-image-turbo-finetune-of-absurd-reality
Graph Convolutional Network
Mentioned in 3 entries
- bi-level-chaotic-fusion-based-graph-convolutional-network-fo
- efficient-and-scalable-granular-ball-graph-coarsening-method
- hierarchical-vision-transformer-enhanced-by-graph-convolutio
ChatGPT Image 2.0
Mentioned in 3 entries
- realistic-results-using-chatgpt-image-20-seedance-20---promp
- replaced-made-with-chatgpt-image-20-seedance-20
- though-the-images-are-very-good-chatgpt-image-20-does-have-t
Kling 3.0
Mentioned in 3 entries
- kling-30-httpslinkscomfyorg4d2amn9
- kling-4k-mode-is-now-supported-in-comfyui-update-comfyui-to-
- same-prompt-3-models-6-camera-moves-we-put-seedance-20-kling
Klein9B
Mentioned in 3 entries
- lciet-and-klein9b-a-quick-fair-comparison-analysis-included
- masterpiece-klein9b-craftsmanship-for-novices
- ultrareal-upscaling-lora-for-klein9b-new-version
Hunyuan
Mentioned in 3 entries
- aigc-video-detection-based-on-the-fusion-of-spatial-frequenc
- pros-making-ai-video-of-real-people-open-source-pipeline-flu
- what-happened-to-hunyuan
kimi-k2.6
Mentioned in 3 entries
- curious-about-ollama-kimi-k26cloud-speed-3-test-runs-779-tok
- ollama-cloud-nerfed-no-more-minimax-m27-or-kimi-k26
- some-models-to-try-with-codex-kimi-k26cloud-with-vision-supp
Flux Klein 9B
Mentioned in 3 entries
- flux-klein-9b---anything-2-real-lora-from-civitaicom
- how-to-retain-lighting-when-remastering-images-local-flux-kl
- rererence-conditioning-reference-image-in-flux-klein-9b---ho
Sapiens2
Mentioned in 3 entries
- human-in-the-loop-made-in-comfyui-with-a-laundry-list-of-too
- sapiens2
- sapiens2-is-the-highest-quality-vit-backbone-that-now-exists
Vera Rubin
Mentioned in 3 entries
- sk-hynix-says-it-has-begun-mass-production-of-the-192gb-soca
- sources-anthropic-has-agreed-to-pay-nscale-45b-over-six-year
- sources-some-of-nvidias-top-customers-have-been-told-that-pr
GPT-imagegen-2
Mentioned in 3 entries
- gpt-imagegen-2-make-5x5-grid-of-dog-photos-where-each-photo-
- presumably-gpt-imagegen-2-aka-chatgpt-images-20-aka-gpt-imag
- this-wasnt-the-case-with-previous-image-generators-but-the-l
Qwen3-VL
Mentioned in 3 entries
Flux 2 Klein 9B
Mentioned in 3 entries
- adonis---general-consistencyupscale-edit-model-for-flux-2-kl
- real-lighting-control-with-flux-2-klein-9b-with-controllight
- update-adonis-post-model---for-adonis---general-consistencyu
MiMo-V2.5-Pro
Mentioned in 3 entries
- mimo-v25-pro-open-sourced
- mimo-v25-pro-when
- xiaomis-mimo-v25-and-mimo-v25-pro-are-both-now-available-in-
Anima 2B
Mentioned in 3 entries
- anima-2b-generation-time
- anima-trainflow-simple-one-page-lora-trainer-for-anima-2b-po
- the-issue-of-repetitive-compositions-in-anima
S4
Mentioned in 3 entries
- advancing-intelligent-sequence-modeling-evolution-trade-offs
- safety-security-and-cognitive-risks-in-state-space-models-a-
- sl-s4wave-self-supervised-learning-of-physiological-waveform
MiniMax Speech 2.8 Turbo
Mentioned in 3 entries
- minimax-speech-28-turbo-is-built-for-voice-agents-that-need-
- see-minimax-speech-28-turbo-in-voice-finder-and-try-the-voic
- we-added-600-new-voices-on-together-ai-introducing-minimax-s
Segment Anything Model 2
Mentioned in 3 entries
- auralsam2-enabling-sam2-hear-through-pyramid-audio-visual-fe
- camosam2-sam2-oriented-prompt-auto-refinement-for-video-camo
- m4-sam-multi-modal-mixture-of-experts-with-memory-augmented-
Amazon Nova Sonic
Mentioned in 3 entries
- build-real-time-voice-streaming-applications-with-amazon-nov
- evaluate-your-amazon-nova-sonic-voice-agent-at-scale-no-micr
- scalable-voice-agent-design-with-amazon-nova-sonic-multi-age
YOLOv12
Mentioned in 3 entries
- coswa-yolov12-scale-invariant-tiny-object-detection-and-segm
- early-detection-of-acute-myeloid-leukemia-aml-using-yolov12-
- using-the-yolov12-model-for-verifying-the-correct-color-sequ
MambaVision
Mentioned in 3 entries
- geomamba-a-geometry-driven-mambavision-framework-and-dataset
- hybrid-quantum-mambavision-a-quantum-enhanced-state-space-mo
- mambarefine-cd-mambavision-with-region-boundary-temporal-ref
Bernini
Mentioned in 3 entries
- anyone-test-bernini-yet-what-vramram-is-already-working
- bernini-released-unified-video-generation-and-editing-model-
- bytedance-bernini-workflow
Anima Base 1.0
Mentioned in 3 entries
- my-steps-and-yours-anima-base-10---qwen-image-edit-2511---wa
- playing-with-anima-base-10-flux2-klein-9b-wan-22-no-audio
- tried-custom-lora-for-anima-base-10-and-its-absolutely-amazi
HunyuanVL
Mentioned in 3 entries
TripoSplat
Mentioned in 3 entries
- getting-started-1-update-comfyui-to-the-latest-version-2-ope
- triposplat-an-open-source-image-to-3d-gaussian-model-from-tr
- triposplat-triposplat-converts-a-single-2d-image-into-high-q
LTX Director
Mentioned in 3 entries
- complex-scene-transitions-with-the-new-ltx-director-and-tran
- local-i2v-finally-feels-less-like-image-wiggle-and-more-like
- ltx-23-ltx-director-is-a-huge-improvement
MedSAM
Mentioned in 3 entries
- enhancing-medsam-with-a-lightweight-box-predictor-for-medica
- medcore-boundary-preserving-medical-core-pruning-for-medsam
- when-adaptation-hurts-connecting-representational-drift-to-o
Bonsai Image 4B
Mentioned in 3 entries
- bonsai-image-4b-a-pair-of-low-bit-diffusion-transformer-depl
- prismml-just-released-binary-and-ternary-bonsai-image-4b-1-b
- testing-the-new-prismml-bonsai-image-4b
Grok Build 0.1
Mentioned in 3 entries
- grok-build-01-is-available-on-vercel-ai-gateway-xais-beta-co
- grok-build-01-is-now-available-for-early-access-in-hermes-ag
- xai-models-now-support-zero-data-retention-zdr-on-openrouter
Graph Convolutional Networks
Mentioned in 3 entries
- impact-detection-in-fall-events-leveraging-spatio-temporal-g
- o-ran-xapps-conflict-prediction-using-graph-convolutional-ne
- particle-competition-and-cooperation-for-robust-graph-convol
SigLIP2
Mentioned in 3 entries
- modernising-the-compressed-domain-video-captioner-a-controll
- technical-highlights-cogvit-vision-encoder---built-with-dual
- turingvit-making-sota-vision-transformers-accessible-to-all
gpt-image-1
Mentioned in 3 entries
- vision-encoder-behavioral-fingerprints-of-image-to-image-gen
- vox-style-match-cut-montage-this-workflow-splits-your-input-
- wheres-the-raccoon-with-the-ham-radio-chatgpt-images-20
ZiT
Mentioned in 3 entries
- a-fully-character-driven-fantasy-story-made-entirely-with-lt
- ernie-vs-qwen-and-zit---big-test
- tested-idg4-against-zits-strong-suit-realistic-portrait
VQ-VAE
Mentioned in 3 entries
- arcvq-vae-a-spherical-vector-quantization-framework-with-arc
- continuous-first-discrete-later-vq-vaes-without-dimensional-
- spatiotemporal-seismic-hazard-assessment-using-vq-vae-and-se
Moebius
Mentioned in 3 entries
- moebius-02b-lightweight-image-inpainting-framework-with-10b-
- my-parallel-agent-side-project-today-was-having-claude-code-
- you-can-try-it-out-on-your-own-images-here-13gb-model-downlo
ESM-2
Mentioned in 3 entries
- orbit-fmib-tracking-order-resolved-epistatic-information-thr
- structural-interpretations-of-protein-language-model-represe
- when-do-cheap-embeddings-beat-protein-language-models-a-theo
Z-Image 6B
Mentioned in 3 entries
- sd-webui-codex-z-image-6b-with-pixel-space-gen-no-vae-thread
- tencent-released-z-image-6b-with-pixel-space-gen-no-vae-1k-r
- testing-z-image-6b-in-comfyui-experimental-pixel-space-workf
Pixal3D
Mentioned in 3 entries
- automated-wrapper-around-pixal3d-comfyui-drop-images-in-a-fo
- i-built-a-free-demo-for-pixal3d-tencent-new-image-to-3d-mode
- pixal3d-generate-high-fidelity-3d-assets-from-a-single-image
GLM5.2
Mentioned in 3 entries
- glm52-7tg-on-4x3090-192gb-on-budget-motherboard-cpu
- orangepi-ai-studio-pro---qwen35-122b-a10b
- should-i-go-back-to-ollama-cloudit-was-great-for-me
Minimax-M3
Mentioned in 3 entries
- b10142
- lwiai-podcast-248---opus-48-mai-anthropic-ipo-minimax-m3
- model-page-httpsollamacomlibraryminimax-m3
Flux.2 Klein 9b
Mentioned in 3 entries
- bare-metal-z-image-turbo---flux2-klein-9b---wan-22
- playing-with-anima-base-10-flux2-klein-9b-wan-22-no-audio
- the-hunt-2-z-image-turbo---flux2-klein-9b---wan-22
DramaBox
Mentioned in 3 entries
- audiocpp-release-05-dramabox-expressive-tts-confucius4-cross
- dramabox---most-expressive-voice-model-ever-based-on-ltx-23
- haha-its-a-dramatic-voice-model
MiniCPM5-1B
Mentioned in 3 entries
- minicpm5-1b-is-an-impressive-release-in-the-1b-class-openbmb
- tool-calls-working-with-minicpm5-1b
- what-openbmb-1b-version-is-this
Gemini Nano
Mentioned in 3 entries
- chromes-4gb-ai-model-isnt-new-but-youre-not-wrong-for-being-
- google-chrome-silently-installs-a-4gb-gemini-nano-model-on-d
- google-says-the-gemini-app-now-offers-personalized-nano-bana
Amazon Nova Forge
Mentioned in 3 entries
- custom-reward-functions-for-multi-turn-reinforcement-learnin
- deploying-multi-turn-rl-infrastructure-for-amazon-nova-on-am
- the-generative-ai-customization-spectrum-from-prompt-enginee
Inception
Mentioned in 3 entries
- classifying-galaxies-in-the-galaxy10-decals-dataset-using-in
- ibis-a-hybrid-inception-bilstm-and-svm-ensemble-for-robust-d
- lightsleepx-a-lightweight-inception-based-dual-modal-network
Qwen3.5-122B-A10B
Mentioned in 3 entries
- orangepi-ai-studio-pro---qwen35-122b-a10b
- release-wintermix-qwen35-122b-a10b-in-native-mlx-an-82-gib-b
- uncensored-multi-model-releases-longcat-flash-lite-sparse-wi
K3
Mentioned in 3 entries
- absurd-claim-the-distilled-model-outperforms-the-originals
- fyi-fireworksaihq-launched-inference-and-training-for-k3-on-
- k3-already-got-in-the-top-5-most-liked-models-of-all-time-on
FLUX 3
Mentioned in 3 entries
- flux-3---real-world-models-towards-multimodal-flow-models-as
- flux-3-is-now-live-on-together-ai-bflais-new-multimodal-mode
- together-serverless-inference-gives-developers-a-managed-pro
Gemma 3 4B
Mentioned in 3 entries
- distilling-self-consistency-into-verbal-confidence-a-pre-reg
- qwen-35-4b-iq2xs-1667-reasoning-performance-from-tensor-leve
- update-on-the-engine-tested-gemma34b-it-actually-reasons-35-
EXAONE 4.5
Mentioned in 3 entries
Graph Attention Networks
Mentioned in 3 entries
- gated-multi-graph-fusion-via-graph-attention-networks-for-al
- probabilistic-salary-prediction-with-graph-attention-network
- spatial-prediction-of-soil-microplastics-and-organic-matter-
MAI
Mentioned in 3 entries
- ainews-microsoft-build-mai-thinking-1-and-mai-family-models
- microsoft-mai-models-coming-soon-to-fireworks-intelligence-y
- sources-microsoft-looking-to-reduce-ai-costs-is-starting-to-
VLA-JEPA
Mentioned in 3 entries
- as-many-of-you-were-interested-in-the-technical-details-of-t
- output-level-regularization-eliminates-the-seed-lottery-in-s
- vla-jepa-just-dropped-in-lerobot-what-makes-this-model-speci
VMamba
Mentioned in 3 entries
- hamsa-scanning-free-vision-state-space-models-via-spectralpu
- scdm-spatial-contextual-disentanglement-mamba-via-differenti
- trust-test-time-refinement-using-uncertainty-guided-ssm-trav
Gaussian Processes
Mentioned in 3 entries
- inducing-permutation-invariant-priors-in-bayesian-optimizati
- laplace-approximation-for-bayesian-tensor-network-kernel-mac
- scalable-bayesian-additive-models-for-stellar-flare-detectio
Echo State Networks
Mentioned in 3 entries
- deep-residual-echo-state-networks-exploring-residual-orthogo
- echo-state-networks-for-time-series-forecasting-hyperparamet
- rethinking-reservoir-pruning-a-dynamical-perspective-for-ech
text-to-image diffusion models
Mentioned in 3 entries
- erase-but-preserve-controllable-removal-of-copyrighted-anima
- forgetting-is-competition-rethinking-unlearning-as-represent
- preserve-and-personalize-personalized-text-to-image-diffusio
CNN-LSTM
Mentioned in 3 entries
- an-improved-cnn-lstm-based-intrusion-detection-system-for-io
- physics-informed-cnn-lstm-for-street-scale-urban-flood-predi
- tsconet-a-two-stage-copula-cnn-lstm-for-uncertainty-aware-sp
TabICLv2
Mentioned in 3 entries
- localized-tabiclv2-scaling-tabular-in-context-learning-throu
- tabiclv2-a-better-faster-scalable-and-open-tabular-foundatio
- tfm-retouche-a-lightweight-input-space-adapter-for-tabular-f
Ling-3.0-tiny
Mentioned in 3 entries
- antlingve-open-sourced-6-base-model-checkpoints-for-ling-30-
- inclusionailing-30-tiny-8b-a13b-moe-hugging-face
- ling-30-tiny-is-a-very-interesting-model-run-on-nvidia-orin-
DeepONets
Mentioned in 3 entries
- deep-set-based-operator-learning-with-uncertainty-quantifica
- fixed-and-adaptive-topological-deeponets-functional-measurem
- learning-unsteady-aneurysm-hemodynamics-with-physics-informe
Swin
Mentioned in 3 entries
- ciunet-a-hybrid-swin-cnn-unet-for-medical-image-segmentation
- swin-meets-efficientnet-lightweight-architectures-for-gan-ba
- swinception----local-attention-meets-convolutions
WeatherNext 3
Mentioned in 3 entries
- introducing-weathernext-3-our-most-advanced-global-weather-a
- weathernext-3-increasing-resolution-and-performance-of-globa
- weathernext-3-will-begin-enhancing-weather-experiences-withi
Qwen3.8-Max-0902
Mentioned in 3 entries
- 1-on-codearena-webdev-leaderboard-qwen38-max-0902-jumps-from
- 1-overall-on-code-arena-and-top-of-the-pareto-frontier-at-5m
- qwen38-max-just-got-upgraded-meet-qwen38-max-0902-24t-parame
Video diffusion models
Mentioned in 3 entries
- dsaquant-denoising-stage-aligned-quantization-aware-training
- exploring-data-free-lora-transferability-for-video-diffusion
- vigor-video-geometry-oriented-reward-for-temporal-generative
Krea2
Mentioned in 3 entries
- chromea-a-chroma-based-lora-for-krea2-training-currently-and
- k2lab-standaloneish-krea2-bbox-style-prompting-and-lora-cont
- why-havent-we-got-an-sota-open-image-edit-model-yet
Conformer
Mentioned in 3 entries
- categorize-early-integrate-late-divergent-processing-strateg
- from-monolingual-to-multilingual-evaluating-mamba-for-asr-in
- neonatal-hypoxic-ischaemic-encephalopathy-classification-fro
Qwen3.8 Flash
Mentioned in 3 entries
Deep Q-Networks
Mentioned in 3 entries
- enhancing-q-value-updates-in-deep-q-learning-via-successor-s
- nimbusguard-a-novel-framework-for-proactive-kubernetes-autos
- revisiting-td-target-aggregation-under-uncertainty-in-q-lear
HuBERT
Mentioned in 3 entries
- babyhubert-multilingual-self-supervised-learning-for-segment
- does-a-prosody-trained-representation-help-beyond-trainable-
- maubert-universal-phonetic-inductive-biases-for-few-shot-aco
Gated DeltaNet-2
Mentioned in 3 entries
- gated-deltanet-2-is-almost-exactly-rwkv-7s-dplr-recurrence-n
- gated-deltanet-2-is-here-new-paper-gated-deltanet-2-decoupli
- gated-deltanet-has-been-one-of-my-favorite-hybrid-attention-
Kimi 2.7
Mentioned in 3 entries
- 52-could-be-better-with-more-rl
- is-glm-52-kimi-27-still-worth-it
- this-is-very-interesting-coinbase-seems-to-have-lowered-thei
Latent Diffusion
Mentioned in 3 entries
- ai-for-cultural-heritage-textiles-fine-tuned-latent-diffusio
- tracking-high-order-evolutions-via-cascading-low-rank-fittin
- vamp-diff-vampprior-latent-diffusion-for-photoplethysmograph
Mamba-3
Mentioned in 3 entries
- beam3r-beams-eye-view-architecture-with-mamba-3-for-implicit
- gated-deltanet-2-is-here-new-paper-gated-deltanet-2-decoupli
- non-commutative-state-tracking-with-input-dependent-low-rank
Vera Rubin NVL72
Mentioned in 3 entries
- how-nvidia-groq-3-lpx-unlocks-ultrafast-interactivity-at-lon
- nvidia-vera-rubin-driving-performance-per-watt-lowest-token-
- spacex-in-partnership-with-nvidia-has-designed-a-space-optim
Paris 2.0
Mentioned in 3 entries
- paris-20-a-decentralized-diffusion-model-for-video-generatio
- were-releasing-paris-20-which-to-our-knowledge-is-the-worlds
- were-selectively-releasing-the-paris-20-weights-and-partneri
Flux 3
Mentioned in 3 entries
- germanys-black-forest-labs-launches-flux-3-and-flux-mimic-it
- may-have-found-the-highest-and-best-use-case-of-flux-3---gen
- will-there-be-flux-3-klein
TD3
Mentioned in 3 entries
- control-of-a-twin-rotor-using-twin-delayed-deep-deterministi
- hybrid-td3-overestimation-bias-analysis-and-stable-policy-op
- noml-noml-hierarchical-td3-anchor-policy-for-flight-control-
BART
Mentioned in 3 entries
- leveraging-bart-to-assess-cs1-c-programming-assignments-usin
- scurank-ranking-multiple-candidate-summaries-with-summary-co
- transformer-models-for-text-summarization-a-comparative-stud
Spiking Vision Transformers
Mentioned in 3 entries
- bsvit-a-burst-spiking-vision-transformer-for-expressive-and-
- morphatt-a-neuromorphic-accelerator-for-efficient-multi-head
- primesvt-an-automated-memory-aware-pruning-framework-with-pr
Detection Transformers
Mentioned in 3 entries
- corrupting-attention-evasion-based-adversarial-attacks-on-en
- enhanced-knowledge-distillation-for-detection-transformer-vi
- fiber-resolved-microstructure-quantification-from-multi-shel
YOLOv11
Mentioned in 3 entries
- comparing-yolov11-and-yolov8-for-instance-segmentation-of-oc
- cross-generation-optimization-of-yolov26-yolov11-and-yolov8-
- performance-analysis-of-yolov11-and-yolov8-for-mixed-traffic
Masked Autoencoders
Mentioned in 3 entries
- do-masked-autoencoders-improve-downhole-prediction-an-empiri
- robust-representation-learning-in-masked-autoencoders
- structure-guided-masked-autoencoders-for-ultra-high-resoluti
WavLM
Mentioned in 3 entries
- advancing-speaker-based-vocal-effort-classification-with-wav
- from-a-to-b-to-a-palindromic-zero-shot-voice-conversion-with
- pase-leveraging-the-phonological-prior-of-wavlm-for-low-hall
ConvNeXt
Mentioned in 3 entries
- attention-enhanced-dual-branch-convnext-bilstm-network-for-s
- new-episode-of-the-information-bottleneck-is-out-this-time-w
- slimmable-convnext-width-adaptive-inference-for-efficient-mu
QNN
Mentioned in 3 entries
- a-stable-aggregation-method-for-quantum-federated-learning
- fisher-information-training-and-bias-in-fourier-regression-m
- image-classification-using-cnn-qnn-hybrid-model-with-optimiz
Qwen2.5-7B
Mentioned in 3 entries
- 28-tps-on-qwen25-7b-across-two-separate-cloud-regions-over-p
- knowledge-distillation-from-large-reasoning-models-to-compac
- probing-latent-colombian-identity-inferences-in-qwen25-7b-wi
Tiny Recursive Models
Mentioned in 3 entries
- laser-low-rank-activation-svd-for-efficient-recursion
- probabilistic-tiny-recursive-model
- tiny-recursive-models-for-solving-the-j2-perturbed-lambert-p
BiomedCLIP
Mentioned in 3 entries
- cadre-stable-parameter-efficient-adaptation-of-medical-visio
- explaining-biomedclip-with-weighted-banzhaf-interactions-sup
- federated-lora-adaptation-of-biomedclip-across-four-internat
YOLO11
Mentioned in 3 entries
- bee-detection-and-tracking-at-hive-entrance-using-yolo11-and
- ultralytics-yolo-evolution-an-overview-of-yolo27-yolo26-yolo
- yolov8-to-yolo11-a-comprehensive-architecture-in-depth-compa
EfficientNet
Mentioned in 3 entries
- attention-guided-efficientnet-architecture-for-precise-crimi
- swin-meets-efficientnet-lightweight-architectures-for-gan-ba
- uncertainty-aware-wildfire-smoke-density-classification-from
Sakana Chat
Mentioned in 3 entries
- product-designer-httpssakanaaicareersproduct-designer-sakana
- sakana-ai-sakana-ai-sakana-chat-sakana-marlin-sakana-fugu
- sakana-ai-sakana-chat-namazu-namazu
Kolmogorov-Arnold Network
Mentioned in 3 entries
- glass-box-deep-learning-for-fdia-detection-in-nonlinear-auto
- interpretable-kolmogorov-arnold-network-with-feature-isolate
- pg-kinn-a-physics-informed-petrov-galerkin-kolmogorov-arnold
TabFM
Mentioned in 3 entries
- introducing-tabfm-a-zero-shot-foundation-model-for-tabular-d
- introducing-tabfm-in-bigquery-predictive-analytics-reimagine
- tabfm-a-zero-shot-foundation-model-for-tabular-data
Restricted Boltzmann Machine
Mentioned in 3 entries
- automatic-extraction-of-road-networks-by-using-teacher-stude
- comparison-of-d-wave-quantum-annealing-and-gibbs-monte-carlo
- hyperbolic-restricted-boltzmann-machine-neural-quantum-state
talkie
Mentioned in 3 entries
- ai-researchers-launch-talkie-a-13b-vintage-language-model-tr
- new-work-with-alecrad-and-davidduvenaud-have-you-ever-dreame
- some-notes-on-talkie-a-new-vintage-language-model-from-a-tea
KAME
Mentioned in 3 entries
- learning-natural-conversational-behavior-in-tandem-speech-to
- two-heads-are-better-than-one-async-knowledge-injection-for-
- were-excited-to-introduce-kame-tandem-architecture-for-enhan
Kolmogorov-Arnold networks
Mentioned in 3 entries
- an-adaptive-framework-for-the-axisymmetric-pulsar-magnetosph
- kolmogorov-arnold-networks-in-nuclear-binding-energy-predict
- quantum-classical-physics-informed-kolmogorov-arnold-network
latent diffusion models
Mentioned in 3 entries
- a-gradient-flow-approach-to-solving-inverse-problems-with-la
- hfi-a-unified-framework-for-training-free-detection-and-impl
- warpi2i-image-warping-for-image-to-image-translation
AdaJEPA
Mentioned in 3 entries
- adajepa-an-adaptive-latent-world-model
- leworld-model-becomes-adaptive-and-meets-model-predictive-co
- warm-take-your-world-model-should-never-stop-learning-introd
ResNet-50
Mentioned in 3 entries
- quantum-inspired-hybrid-neural-networks-for-neural-decoding-
- resnet-50-with-class-reweighting-and-anatomy-guided-temporal
- skin-lesion-classification-based-on-resnet-50-enhanced-with-
ResNet50
Mentioned in 3 entries
- a-comparative-study-of-gan-based-deep-learning-models-for-pn
- classification-of-disease-from-lungs-x-ray-images-using-vgg1
- external-validation-of-deep-learning-models-for-bi-rads-brea
Talkie
Mentioned in 3 entries
- amazing-its-like-talking-to-an-ai-from-the-past
- announcing-talkie-a-new-open-weight-historical-llm-we-traine
- test-of-time-llm-from-some-legendspretty-fun
DreamBooth
Mentioned in 3 entries
- iddm-identity-decoupled-personalized-diffusion-models-with-a
- metacloak-jpeg-jpeg-robust-adversarial-perturbation-for-prev
- regain-restoring-subject-fidelity-in-personalization-on-synt
pi_0
Mentioned in 3 entries
- flowhijack-a-dynamics-aware-backdoor-attack-on-flow-matching
- refactor-vla-unsupervised-library-learning-of-typed-motor-pr
- reinforcement-fine-tuning-of-flow-matching-policies-for-visi
Sakana Namazu
Mentioned in 3 entries
- introducing-sakana-namazu-an-llm-api-with-japanese-vibes-tod
- member-of-technical-staff-llm-development-httpssakanaaicaree
- sakana-namazu-an-llm-api-with-japanese-vibes-built-for-japan
TabICL
Mentioned in 3 entries
- distributional-regression-with-tabular-foundation-models-eva
- scoringbench-a-benchmark-for-evaluating-tabular-foundation-m
- tabiclv2-a-better-faster-scalable-and-open-tabular-foundatio
GPT-5.3 Instant
Mentioned in 3 entries
- gpt-55-instant-is-rolling-out-now-in-chatgpt
- gpt-55-instant-is-starting-to-roll-out-in-chatgpt
- openais-gpt-55-instant-the-good-the-bad-and-the-insane
Qwen3.5 0.8B
Mentioned in 3 entries
- procedural-skill-sft-across-capacity-tiers-a-w-shaped-pre-sf
- qwen35-08b-on-cpu
- where-are-small-models-like-qwen3-06b-and-qwen35-08b-used-hu
Fugu-Ultra v1.1
Mentioned in 3 entries
- announcing-fugu-ultra-v11-weve-been-thrilled-by-the-receptio
- announcing-the-claude-code-compatible-interface-for-our-new-
- our-team-just-shipped-fugu-ultra-v11-by-dynamically-orchestr
GLM5
Mentioned in 3 entries
- after-switching-into-the-new-directory-cd-lc-docs-im-going-t
- b11352
- open-harness-deployed-agents-if-you-wanna-use-claude-glm5-an
MAI-Cyber-1-Flash
Mentioned in 3 entries
- microsoft-introduces-mai-cyber-1-flash-an-ai-model-trained-f
- microsoft-says-mai-cyber-1-flash-and-mdash-its-vulnerability
- microsofts-introduces-its-first-agent-powered-cybersecurity-
text-embedding-3-small
Mentioned in 3 entries
- correction-openai-are-not-deprecating-that-model-the-announc
- openai-is-shutting-down-text-embedding-3-small-i-strongly-be
- thank-you-for-flagging-this-jeff-this-was-a-mistake-we-are-n
Ling-3.0-flash-VL
Mentioned in 3 entries
- inclusionailing-30-flash-vl-hugging-face
- ling-30-flash-vl-built-on-ling-30-flash-with-visual-understa
- model-add-ling-30-vl-support-by-aetherbird-pull-request-2915
K2 Horizon
Mentioned in 3 entries
- ama-about-k2-horizon-meet-our-team-from-ifm
- introducing-k2-horizon-a-connected-fleet-of-six-foundation-m
- k2-horizon-lineup-is-out-on-aa-and-once-again-aa-plots-are-m
Gemini 2
Mentioned in 3 entries
- buckle-up-google-is-set-to-remake-search-with-agentic-ai-in-
- httpsxcomchrisfirststatus2056797606509158681s20
- multi-token-prediction-mtp-for-llamacpp-running-gemma4-local
Watermelon
Mentioned in 3 entries
- after-meta-avocado-we-get-watermelon-due-in-november
- docs-meta-plans-to-launch-its-version-of-openclaw-codenamed-
- sources-alexandr-wang-said-metas-model-currently-in-training
Toto 2.0
Mentioned in 3 entries
- are-scaling-laws-finally-working-for-time-series-foundation-
- toto-20-is-here-datadog-ais-5-open-weights-forecasting-model
- toto-20-time-series-forecasting-enters-the-scaling-era
Spark
Mentioned in 3 entries
- ainews-google-io-2026-gemini-35-flash-omni-nanobanana-for-vi
- full-interview-with-sundar-is-now-live-httpsxcomrowancheungs
- one-of-the-things-the-pelican-benchmark-is-still-useful-for-
Full Self-Driving
Mentioned in 3 entries
- i-tried-tesla-full-self-driving-over-the-weekend-its-magical
- the-human-perceived-rgb-is-image-1-and-the-tesla-ai-photon-c
- try-tesla-fsd
Claude 5.5
Mentioned in 3 entries
- both-are-great-models-and-neither-wins-everywhere-i-use-both
- i-asked-55-thinking-to-generate-an-image-of-its-forbidden-wi
- wait-if-most-people-think-55-is-better-than-47-i-assume-that
SL2T
Mentioned in 3 entries
- google-debuts-sl2t-an-ai-model-thats-designed-to-understand-
- google-deepmind-launches-sl2t-a-multilingual-sign-language-t
- putting-sign-language-ai-into-users-hands
Qwen3-Coder
Mentioned in 3 entries
Claude Sonnet 4
Mentioned in 3 entries
- banger-paper-from-stanford-and-together-ai-they-show-why-it-
- llm-driven-automl-for-cross-lingual-handwritten-ocr-closed-l
- study-shows-ai-chooses-nuclear-war-in-crisis
Daybreak Blue
Mentioned in 3 entries
- accelerate-cyber-defense-with-openai-and-aws-daybreak-red-da
- codex-security-cloud-is-getting-a-major-upgrade-with-access-
- openai-unveils-two-daybreak-tiers-daybreak-blue-which-provid
GPT 5.6 Terra
Mentioned in 3 entries
- ainews-openai-launches-gpt-56-solterraluna-codex-becomes-cha
- congrats-to-openai-on-the-highly-anticipated-launch-of-gpt-5
- gpt-56-sol-luna-and-terra-now-available-on-ai-gateway
Parcae
Mentioned in 3 entries
- parcae-doing-more-with-fewer-parameters-using-stable-looped-
- the-edge-inference-implication-memory-not-compute-is-the-bin
- what-if-you-could-get-13b-transformer-quality-from-a-770m-mo
Cosmos3
Mentioned in 3 entries
- measuring-language-transfer-in-robot-policies-adding-greek-t
- sota-imagegen-locally-nvidia-cosmos364b-int4-quants-cudamlx
- tacsushi-tactile-grounded-world-action-modeling-for-dexterou
Gemini 3.5 Pro
Mentioned in 3 entries
- a-hot-tip-for-googledeepmind-open-source-gemini-35-pro-and-b
- by-the-end-of-the-year-we-should-have-gpt-6-fable-55-gemini-
- is-gemini-35-pro-being-export-controlled-because-if-not
Gemini 3.1 Flash
Mentioned in 3 entries
- gemini-31-flash-tts-is-rolling-out-in-google-vids-and-is-ava
- learning-causality-from-internet-videos-in-latent-space-firs
- nano-banana-2-lite-gemini-31-flash-lite-image-now-on-ai-gate
DeepSeek v4 Flash 0731
Mentioned in 3 entries
- benchmark-llamacpp-batchubatch-impacts-on-pp-and-tg
- deepseek-v4-flash-0731-locally-on-cpu
- validate-your-local-llm-advertised-kv-cache-against-real-pre
Gemini 3.1 Flash-Lite
Mentioned in 3 entries
- gemini-31-flash-lite-is-now-generally-available-on-gemini-en
- i-switched-my-httpsagentdatasetteio-instance-to-luna-it-was-
- llm-gemini-031
Krea2 Turbo
Mentioned in 3 entries
- krea2-turbo-distill-2-step-lora---new-checkpoint-released-ch
- krea2-turbo-distill-4-step-lora---new-checkpoint-chk26k-rele
- krea2-turbo-distill-4-step-lora-not-a-new-checkpoint-yet-to-
Daybreak
Mentioned in 3 entries
- openai-gpt-56-cyber-daybreak-expansion
- openai-just-launched-a-cybersecurity-model-that-answers-95-o
- openai-says-it-will-give-its-ai-cyber-defense-system-daybrea
Genie
Mentioned in 3 entries
- how-anthropics-new-results-post-would-read-without-the-pr-cl
- i-am-so-tired-of-the-pr
- worldmark-a-unified-benchmark-suite-for-interactive-video-wo
TabPFN-3
Mentioned in 3 entries
- tabpfn-3-just-released-a-pre-trained-tabular-foundation-mode
- tabpfn-3-technical-report
- tabpfn-35-technical-report
OpenAI o1
Mentioned in 3 entries
- lead-length-efficient-adaptive-and-dynamic-reasoning-for-lar
- smartthinker-progressive-chain-of-thought-length-calibration
- when-to-retrieve-during-reasoning-adaptive-retrieval-for-lar
Llama 3.1
Mentioned in 3 entries
- me-arguing-with-an-ai-bot-who-just-posted-something-on-this-
- open-tq-metal-fused-compressed-domain-attention-for-long-con
- see-eg-httpswwwscmpcomtechtech-trendsarticle3271858ai-race-a
Qwen-Image-3.0-Pro
Mentioned in 3 entries
- lets-create-with-qwen-image-30-pro-on-fal
- qwen-image-30-pro-has-made-a-massive-leap-from-the-previous-
- qwen-image-30-pro-is-live-on-qwen-cloud-now-try-it-out-https
Gemma 31B
Mentioned in 3 entries
- cross-generational-transfer-of-adversarial-attacks-reveals-n
- gemma4-31b-vs-qwen38-27b---why-the-huge-difference-in-benchm
- thinking-about-grabbing-an-rtx-2000-ada-16gb-to-add-to-my-ga
Claude Cowork
Mentioned in 3 entries
- anthropic-brings-cowork-directly-inside-claudes-chat-interfa
- liteparse-should-be-the-default-document-parser-you-use-with
- pwc-expands-anthropic-alliance-will-train-30000-staff-on-cla
Muse-Glimmer
Mentioned in 3 entries
- i-asked-deepseek-v4-flash-to-work-with-muse-glimmer-for-visi
- liked-muse-so-i-cut-the-30b-model-in-half-by-width-distilled
- observations-on-muse-glimmer-reasoning-traces-being-noticeab
DeepSeek-V4 Pro
Mentioned in 3 entries
- deepseek-v4-pro-now-available-on-together-ai
- last-week-we-announced-deepseek-v4-today-were-sharing-a-clos
- swe-rebench-multilingual-update-go-java-python-rust-ts-evalu
Llama-3.1-8B-Instruct
Mentioned in 3 entries
- an-effective-rank-audit-of-alignment-induced-activation-shif
- ggml-zendnn-add-q80-quantization-support-by-z-sachin-pull-re
- the-shape-of-wisdom-decision-trajectories-in-language-models
Qwen4Exp
Mentioned in 3 entries
Gemma4-12B
Mentioned in 3 entries
- good-guy-deepseek-gives-us-accelerated-models-the-most-inter
- httpshuggingfacecocollectionsdeepseek-aideepspec
- nanbeige42-3b-im-not-impressed
Qwen3.6 35B
Mentioned in 3 entries
- qwen36-27b-35b-on-vllm-single-r9700-gfx1201
- qwen36-35b-2-min-vs-muse-glimmer-30b-4-min-on-custom-llamacp
- qwen36-35b-q2xxs-being-gpu-poor-in-2026-is-not-so-bad
Broad Learning System
Mentioned in 3 entries
- eca-bls-an-efficient-complex-augmented-broad-learning-system
- ifw-bls-dual-robust-broad-learning-system-with-intuitionisti
- robust-broad-learning-system-with-wave-loss-for-classificati
Gemini 3.5 Flash Lite
Mentioned in 3 entries
- llm-gemini-033
- openwiki-now-supports-gemini-ai-studio-enterprise-vertex-ai-
- we-benchmarked-gemini-36-flash-and-gemini-35-flash-lite-on-d
Gemini 3.1 Flash Lite
Mentioned in 3 entries
- deepseek-v4-flash-is-90-cheaper-than-gpt-54-mini-and-70-chea
- nano-banana-2-lite
- we-benchmarked-gemini-36-flash-and-gemini-35-flash-lite-on-d
Gemini 3.5 Live Translate
Mentioned in 3 entries
- googles-gemini-35-live-translate-enables-realistic-translati
- the-latest-ai-news-we-announced-in-june-2026
- today-we-released-gemini-35-live-translate-our-latest-audio-
Claude Haiku
Mentioned in 3 entries
- claude-haiku-is-my-current-least-favorite-model---it-halluci
- claudes-voice-mode-is-now-available-for-opus-and-sonnet
- prompt-optimization-is-a-coin-flip-diagnosing-when-it-helps-
BabyLMs
Mentioned in 3 entries
- child-directed-speech-facilitates-production-not-comprehensi
- cross-linguistic-effects-in-bilingual-phoneme-babylms
- feeding-babylms-macaroni-code-switching-curricula-cause-cros
Gemini 2.5
Mentioned in 3 entries
- law-professors-wrote-questions-they-were-asked-during-office
- what-would-it-take-for-the-frontier-labs-to-open-the-weights
- when-does-data-augmentation-help-evaluating-llm-and-back-tra
Sydney
Mentioned in 3 entries
- from-this-perspective-gemini-is-also-worse-than-the-original
- if-llama-4-didnt-fail-if-microsoft-had-pulled-sydney-after-t
- when-you-are-talking-to-an-llm-you-are-speaking-to-a-synthes
Claude 4.7
Mentioned in 3 entries
- 6-give-claude-a-way-to-verify-its-work-finally-make-sure-cla
- i-have-found-47-great-for-design-reverted-back-to-46-extende
- in-claude-code-the-default-effort-is-now-xhigh-a-new-level-b
Mistral OCR
Mentioned in 3 entries
- parsebench-is-now-also-available-on-papers-with-code-find-it
- we-benchmarked-mistral-ocr-against-other-frontier-and-open-w
- weve-provided-some-updated-results-on-mistral-ocr-that-make-
Nemotron 3.5 ASR
Mentioned in 3 entries
- how-to-fine-tune-nemotron-35-asr-for-your-language-domain-or
- introducing-two-nvidia-nemotron-models-on-together-ai-nemotr
- nemotron-35-asr-is-built-for-streaming-multilingual-speech-r
Nemotron-Labs-Diffusion
Mentioned in 3 entries
- lwiai-podcast-252---gpt-56-grok-45-nemotron-labs-diffusion-a
- nemotron-labs-diffusion-a-tri-mode-language-model-unifying-a
- were-releasing-nemotron-labs-diffusion---the-first-tri-mode-
Cursor
Mentioned in 3 entries
- new-signadot-skill-lets-claude-code-codex-and-cursor-validat
- qodo-just-launched-the-agentic-toolbox-it-gives-claude-code-
- we-just-launched-comfy-mcp-in-public-beta-the-first-mcp-buil
TrOCR
Mentioned in 3 entries
- adapting-trocr-for-printed-tigrinya-text-recognition-word-aw
- devanagari-handwritten-character-recognition-using-trocr-a-t
- trocr-for-medieval-htr-a-systematic-ablation-study-with-cros
Qwen-Image-Flash
Mentioned in 3 entries
- nvidia-releases-qwen-image-flash
- qwen-image-flash-beyond-objective-design
- qwen-image-flash-papers---qwen-image-2-distilled-turbo
DeepSeek V4 0731
Mentioned in 3 entries
- deepseek-v4-0731---qwen-38-flash---glm-53-flash-and-back-aga
- oldpro-plan-53flash-v41flash-0731-confusion-regarding-usages
- we-quantized-deepseek-v4-0731-and-benchmarked-it-against-pop
Mistral Medium 3.5
Mentioned in 3 entries
- command-a-sets-a-new-high-for-coheres-machine-translation-ca
- introducing-remote-agents-in-vibe-and-mistral-medium-35-you-
- the-story-gets-bigger-beyond-europe-command-a-makes-major-ga
GPT-5.1
Mentioned in 3 entries
- embodied-gpt-51-evidence-of-a-world-model
- openai-says-its-models-starting-with-gpt-51-increasingly-men
- we-published-a-new-ai-safety-study-shopping-agents-fall-for-
Llama-3.1-8B
Mentioned in 3 entries
- arithmetic-in-the-wild-llama-uses-base-10-addition-to-reason
- benchmarking-linguistic-adaptation-in-comparable-sized-llms-
- small-llms-pruning-vs-training-from-scratch
Gemini Embedding 2
Mentioned in 3 entries
- built-a-bird-search-app-to-show-off-pinecones-full-text-sear
- gemini-embedding-2-a-native-multimodal-embedding-model-from-
- the-aidotengineer-europe-conference-last-week-was-a-blast-fu
Qwen 3.6-35B-A3B
Mentioned in 3 entries
- heres-qwen-36-35b-a3b-vs-claude-opus-47-for-generate-an-svg-
- qwen-3635b-on-24-vram-gpu
- vlm-performanceqwen36-is-natively-multimodal-and-qwen36-35b-
Claude Mythos 5.1
Mentioned in 3 entries
- anthropic-launches-claude-fable-51-after-inking-35b-cloud-de
- claude-fable-51-made-me-a-really-nice-animated-pelican
- four-major-ai-models-suffer-rare-overlapping-downtime
Ling-3.0
Mentioned in 3 entries
- ling-30-bailingmoe3-lands-in-llamacpp-mainline---quick-bench
- ling-30-released-all-6-base-checkpoints-2-sizes-3-stages
- llamacpp-pr-26608-ling-30-support-unmerged
Qwen 3.6 35b
Mentioned in 3 entries
- kat-coder-25-dev-do-yourself-a-favor-and-try-it
- qwen-38-35ba3b-wen
- what-are-the-best-models-you-can-run-on-your-nvidiaai-dgx-sp
Spark-X2.5-4B
Mentioned in 3 entries
- minicpm5-2b-vs-spark-x25-4b--64-thinking-x15-speed-while-kee
- model-support-for-spark25forcausallm-implementation-by-knigh
- new-model-spark-x25-4b-spark-x25-17b
Qwen-3.8-Flash-Next
Mentioned in 3 entries
- exllamav3-comfortably-beats-llamacpp-running-cpu-offloaded-q
- exllamav3-recent-updates-cpu-offload-glm-53-flash-qwen38-fla
- qwen-38-flash-next-on-1x-rtx-5090-tg50-ts-pp2300-ts---with-f
Qwen 2.5 7B
Mentioned in 3 entries
- benchmarked-qwen-3-8b-llama-31-8b-qwen-25-7b-vs-gpt-4o-on-m4
- testing-qwen-25-7b-for-geopolitical-multi-agent-simulations-
- whats-currently-the-smartest-llm-to-use-on-8gb-vram-and-16-r
Deepseek V4 Pro
Mentioned in 3 entries
- cancelling-my-subscription-also-it-was-great
- deepseek-v4-pro-ollama-cloud-is-not-working-compare-to-openc
- is-deepseek-v4-pro-working-in-ollama
Qwen3.8-flash-next
Mentioned in 3 entries
- 2rtx-3090-epyc-box-running-qwen38-flash-next-at-38-toks
- qwen38-flash-next-sees-corruption-everywhere
- two-96-gb-ascend-cards-crun-qwen38-flash-next-hardware-notes
Qwen Flash
Mentioned in 3 entries
- if-one-hour-of-ai-is-costing-me-012-is-paying-for-frontier-a
- two-local-qwen-38-27b-unsloth-q6-and-qwen-flash-next-strata-
- vllm-p2p-driver-hack-qwen-38-27b-vs-llamacpp-qwen-flash-next
Qwen 3.7 Max
Mentioned in 3 entries
- qwen-37-max-is-now-supported-in-hermes-agent
- qwen-37-max-now-available-on-vercel-ai-gateway
- qwen-37-max-one-shots-a-5-slide-deck-with-ai-generated-image
BLIP
Mentioned in 3 entries
- a-paragraph-is-worth-a-thousand-captions-rethinking-text-sup
- lifelong-learning-in-vision-language-models-enhanced-ewc-wit
- reassessing-global-gradient-norm-imbalance-in-blip-fine-tuni
Multi-Layer Perceptrons
Mentioned in 3 entries
- benefits-of-low-cost-bio-inspiration-in-the-age-of-overparam
- is-the-statistical-advantage-worth-the-cost-an-empirical-com
- scalability-analysis-of-distributed-kolmogorov-arnold-networ
Llama 2
Mentioned in 3 entries
- b11033
- eladgil-bs-attention-was-born-in-montral-pytorch-in-nyc-alph
- is-this-real-qwen3627b-with-128k-context-fit-in-24gb-vram
Gemini 3.5 Flash-Lite
Mentioned in 3 entries
- google-says-free-gemini-users-will-be-limited-to-the-35-flas
- introducing-gemini-36-flash-35-flash-lite-and-35-flash-cyber
- its-been-a-busy-couple-of-weeks-icymi-heres-the-recap-gemini
Grok 4
Mentioned in 3 entries
- elon-musk-turns-55-today-here-are-55-milestones-age-54-world
- the-new-grok-comes-in-below-the-latest-chinese-open-weights-
- try-grok-45
Qwen3.8 Flash-Next
Mentioned in 3 entries
- qwengram-08b-i-transferred-qwen38-flash-nexts-n-gram-memory-
- what-to-run-at-128gb-vram
- which-current-local-models-that-can-run-within-128gb-generat
ColPali
Mentioned in 3 entries
- adamerge-tuning-free-patch-compression-for-multi-vector-visu
- letzcross-a-cross-lingual-page-level-benchmark-for-multimoda
- structural-anchor-pruning-training-free-multi-vector-compres
Ling 3.0
Mentioned in 3 entries
Qwen3.8-27B-NVFP4
Mentioned in 3 entries
- i-spent-3-weeks-testing-local-qwen38-on-the-new-low-latency-
- qwen38-27b-nvfp4-1m-context-so-far-so-good
- we-pointed-meta-zenith-at-one-task-make-qwen38-27b-nvfp4-fas
Deepseek V4.1 Flash
Mentioned in 3 entries
- deepseek-v41-flash-is-748b-not-552b
- deepseek-v41-flash-release-video-made-with-deepseek-v41-flas
- r9v-update-created-and-adopted-kva-projections-based-on-deep
Qwen-Image 2.1
Mentioned in 3 entries
- i-added-qwen-image-21-lora-support-to-tensorsharp-gguf-local
- i-built-a-free-local-qwen-image-21-app-for-apple-silicon-mac
- qwen-developers-on-x-qwen-image-21-is-going-open-source
Bonsai 2 27B
Mentioned in 3 entries
- prismml-launches-bonsai-2-27b-a-high-intelligence-ai-model-s
- prismml-releases-bonsai-2-27b-which-compresses-alibabas-qwen
- this-is-insane-my-bonsai-2-27b-build-is-trending-on-hugging-
DFM Mimir 1B
Mentioned in 3 entries
Qwen3.8-Omni-Flash
Mentioned in 3 entries
- meet-qwen38-omni-flash-qwens-first-omni-modal-model-built-ar
- qwen38-omni-towards-native-omni-modal-agents
- the-era-of-omni-agents-is-upon-us-this-is-a-great-report-by-
Kev 4B
Mentioned in 3 entries
- jev-vs-kev-open-source-jev-alternative-tested-side-by-side
- kev-4b-topped-out-in-every-tetris-game-i-ran-mica-v01-4b-cle
- kev-tiny-jev-like-decision-models-08b4b9b-on-qwen35-you-can-
Hemmingway-1
Mentioned in 3 entries
- hemmingway-1-an-ai-that-writes-like-a-person-qwen38-27b-fine
- hemmingway-1-an-apache-20-27b-creative-writing-fine-tune-qwe
- we-made-a-27b-model-for-creative-writing-performs-as-good-as
Julia-1
Mentioned in 3 entries
- cool-to-see-classification-models-resurge-in-the-era-of-agen
- llama-server-add-v1systemone-api-models-laya-julia-1-lev-ope
- supersoniclabsjulia-1-hugging-face
vision transformers
Mentioned in 3 entries
- do-vision-model-see-like-the-brain-a-comparison-across-eeg-e
- routing-probes-can-improve-without-new-information-an-exact-
- vg-tie-an-interpretable-tabular-to-image-encoding-method-bas
MiMo-V2.6-Flash
Mentioned in 3 entries
- mimo-v26-both-pro-and-flash-is-a-benchmaxxed-scam
- mimo-v26-flash-on-vllm-fixes-for-empty-responses-with-thinki
- two-300b-moe-models-each-on-one-128-gb-mini-pc-amd-strix-hal
Gemini 3.8 Flash-Lite TTS
Mentioned in 3 entries
- check-out-this-weeks-updates-and-releases-gemini-38-flash-tt
- google-launches-two-benchmark-topping-speech-generation-mode
- were-launching-gemini-38-flash-tts-and-gemini-38-flash-lite-
VeriLoop E2
Mentioned in 3 entries
- veriloop-e2-gguf-release-bf16-iq1m-5011-gib-1679-gib-with-03
- veriloop-e2-is-now-open-on-hugging-face-curious-what-ollamag
- we-released-veriloop-e2-27b-apache-20-the-design-question-be
Nimble
Mentioned in 3 entries
- 3-models-are-now-available-on-ollama-for-use-with-v1systemon
- ollama-now-supports-jev-like-decision-models-all-locally-use
- ollama-tested-tev1-nimble-and-laya-on-my-pacman-arena-game-s
Praxis-1
Mentioned in 3 entries
- announcing-our-first-open-weights-world-action-model-praxis-
- breakthroughs-on-every-frontier-i-cant-wait-to-share-more-on
- today-were-announcing-praxis-1-an-open-weight-world-action-m
Griffin
Mentioned in 3 entries
- remote-work-is-so-cooked-if-your-job-can-be-done-on-the-othe
- tavus-unveils-griffin-the-first-human-interaction-model-whic
- this-is-impressive-face-to-face-conversation-is-where-ai-int
Deep Operator Network
Mentioned in 2 entries
- neural-scaling-laws-of-deep-relu-and-deep-operator-network-a
- probabilistic-predictions-of-process-induced-deformation-in-
anima-preview2
Mentioned in 2 entries
- light-novel-book-illustrations-using-anima-preview2-and-anim
- light-novel-style-book-illustrations-with-anima-preview2
OmniVoice
Mentioned in 2 entries
- omnivoice-multilingual-local-tts-with-600-languages-voice-cl
- omnivoice-towards-omnilingual-zero-shot-text-to-speech-with-
HappyHorse-1.0
Mentioned in 2 entries
Qwen 8B
Mentioned in 2 entries
- ideogram-tip-use-generate-text-node-to-make-json-with-qwen-8
- recommended-model-for-a-4060ti-8gb-and-16gb-ram
Wan 2.1
Mentioned in 2 entries
- creating-unique-visual-styles-for-your-videos-with-wan-21
- experimenting-with-a-cinematic-look-using-wan-21-and-flux-wh
Wan 2.7
Mentioned in 2 entries
- happyhorse-is-from-alibaba-ath-not-grok-veo-32-wan-27-seedan
- this-april-we-shipped-14-new-models-seedance-20-gpt-image-2-
BitNet b1.58
Mentioned in 2 entries
- nativeternary-a-self-delimiting-binary-encoding-with-unary-r
- was-bitnet-a-dead-end-what-happened-to-ternary-llms
Amazon Nova Micro
Mentioned in 2 entries
- cost-efficient-custom-text-to-sql-using-amazon-nova-micro-an
- customize-amazon-nova-models-with-amazon-bedrock-fine-tuning
Amazon Nova Lite
Mentioned in 2 entries
- aderant-builds-intelligent-ticket-triage-with-amazon-nova
- customize-amazon-nova-models-with-amazon-bedrock-fine-tuning
Conditional Neural Processes
Mentioned in 2 entries
- characterizing-the-representational-capacity-of-neural-proce
- exploring-temporal-representation-in-neural-processes-for-mu
pi^3
Mentioned in 2 entries
- does-the-vggt-family-need-all-its-layers
- zipmap-linear-time-stateful-3d-reconstruction-via-test-time-
FSD v14.3
Mentioned in 2 entries
- fsd-v1431-review-after-10-drives-and-many-hours-here-are-my-
- tesla-v143-self-driving-review-the-point-releases-will-bring
Hermes 3
Mentioned in 2 entries
- hermes-will-soon-be-a-serious-creative-tool-more-on-this-soo
- one-hermes-install-later-things-just-click
BiGRU
Mentioned in 2 entries
- enhancing-bigru-with-a-kan-block-for-legal-document-classifi
- hybrid-resnet-1d-bigru-with-multi-head-attention-for-cyberat
SWE-1.6 Fast
Mentioned in 2 entries
- intelligence-at-1000-tokens-per-second-right-in-your-termina
- ok-cognition-swe-16-fast-is-better-than-expected-had-it-buil
Tesla FSD
Mentioned in 2 entries
- a-93-year-old-lady-filmed-her-tesla-fsd-taking-her-to-church
- grok-and-tesla-fsd-are-literally-saving-lives-here-are-the-t
Lyria
Mentioned in 2 entries
- google-wont-just-admit-its-feeding-youtube-creators-to-its-m
- lyria-by-googledeepmind-is-next-level-it-did-average-turkish
Grok-4.20
Mentioned in 2 entries
FLUX Klein
Mentioned in 2 entries
PINN
Mentioned in 2 entries
- lnn-pinn-a-unified-physics-only-training-framework-with-liqu
- mechanical-analysis-of-parachute-suspension-line-deployment-
Qwen 3.5 Plus
Mentioned in 2 entries
Nova 2 Lite
Mentioned in 2 entries
ACE-Step 1.5 XL
Mentioned in 2 entries
ARC-AGI-3
Mentioned in 2 entries
- jensen-huang-says-that-agi-has-been-achieved-i-strongly-disa
- join-the-arc-prize-team----help-us-build-arc-agi-4-and-arc-a
Claude Managed Agents
Mentioned in 2 entries
- deep-agents-deploy-an-open-alternative-to-claude-managed-age
- if-you-prefer-blog-format-httpsbloglangchaincomdeep-agents-d
SMPL
Mentioned in 2 entries
- etch-x-robustify-expressive-body-fitting-to-clothed-humans-w
- more-motion-is-not-always-better-motion-corpus-composition-g
LTX-2.3-22b-IC-LoRA-Outpaint
Mentioned in 2 entries
LTX-Video 2.3
Mentioned in 2 entries
Qwen-Image-Edit
Mentioned in 2 entries
ArcFace
Mentioned in 2 entries
- arcface-embeddings-quantized-to-16-bit-pgvector-halfvec-d
- sparsity-inducing-divergence-losses-for-biometric-verificati
AnimateDiff
Mentioned in 2 entries
- built-a-open-source-local-music-video-generator-using-sdxl-a
- local-ai-tools-for-turning-drawings-into-videos-animatediff-
Z-Image-Base
Mentioned in 2 entries
Llama 4 Maverick
Mentioned in 2 entries
TinyML
Mentioned in 2 entries
- fully-autonomous-z-score-based-tinyml-anomaly-detection-on-r
- webserial-vision-training-for-microcontrollers-a-browser-bas
LTX-2.3 Distilled
Mentioned in 2 entries
Qwen3.5 27B
Mentioned in 2 entries
- qwen35-27b-uncensored-heretic-native-mtp-preserved-is-out-no
- run-unsloth-qwen35-ggufs-or-your-finetuned-version-in-ollama
Qwen3.5 122B-A10B
Mentioned in 2 entries
LTX Video
Mentioned in 2 entries
DeepSeek v3.2
Mentioned in 2 entries
GNNs
Mentioned in 2 entries
- a-unified-graph-language-model-for-multi-domain-multi-task-g
- n-ama-announcement-max-welling-vaes-gnns-ai4science-cuspai
Mantis
Mentioned in 2 entries
- mantis-a-foundation-model-for-mechanistic-disease-forecastin
- mantis-lightweight-foundation-model-for-time-series-classifi
Pony
Mentioned in 2 entries
- i-got-tired-of-writing-massive-prompts-so-i-built-a-local-ra
- pony-faceid-for-same-character-output-keeps-coming-back-as-a
ERNIE Turbo
Mentioned in 2 entries
- complex-weird-prompt-test-ernie-turbo-flux2-klein-4b-z-image
- side-by-side-comparison-of-qwen-image-ernie-baseturbo-and-fl
Anima Preview 3
Mentioned in 2 entries
LTX-Video
Mentioned in 2 entries
Gemma4 26B
Mentioned in 2 entries
- gem16---custom-engine-for-gemma4-12b-26b-on-blackwell-16gb-g
- openclaw-with-gemma4-26b-extremely-slow-and-forget-stuff
Grok 4.3 Beta
Mentioned in 2 entries
- grok-43-beta-browser-os-test-result-gta-clone-and-voice-cont
- grok-43-beta-is-insanely-good-and-now-comes-with-powerful-ne
FinBERT
Mentioned in 2 entries
- domain-fine-tuning-finbert-on-finnish-histopathological-repo
- ra-finbert-rule-aware-lora-adaptation-for-low-resource-finan
ZImage Base
Mentioned in 2 entries
Ltx 2.3
Mentioned in 2 entries
Llama 3.2
Mentioned in 2 entries
- best-ollama-model-for-n8n-workflows-rag-file-handling-reason
- the-capacity-of-thought-benchmarking-llama-32-in-semantic-fm
Flux Dev.1
Mentioned in 2 entries
Flux2.Klein
Mentioned in 2 entries
- flux2klein-tile-upscaler-node-basically-usdu-with-quality-of
- pantomime-facial-expression-sprite-generator-using-flux2klei
Klein 4B
Mentioned in 2 entries
- difference-between-klein-4b-and-klein-9b-is-sooo-big
- same-prompt-on-open-source-models-z-image-base-distilled-kle
Segment Anything Model 3
Mentioned in 2 entries
- activesam-fast-and-accurate-open-vocabulary-semantic-segment
- sam3-i-segment-anything-with-instructions
DenseNet121
Mentioned in 2 entries
- cbam-enhanced-densenet121-for-multi-class-chest-x-ray-classi
- external-validation-of-deep-learning-models-for-bi-rads-brea
Masked Autoencoder
Mentioned in 2 entries
- assessing-the-potential-of-masked-autoencoder-foundation-mod
- pacx-mae-physiology-augmented-chest-x-ray-masked-autoencoder
TranslateGemma
Mentioned in 2 entries
- follow-up-on-the-translategemma-subtitle-benchmark-human-rev
- we-benchmarked-translategemma-against-5-other-llms-on-subtit
FINER
Mentioned in 2 entries
- fire-frequency-reparameterization-as-a-preconditioner-for-pe
- fm-siren-fm-finer-implicit-neural-representation-using-nyqui
Random Forests
Mentioned in 2 entries
- adaptive-msd-splitting-enhancing-c45-and-random-forests-for-
- cross-paradigm-knowledge-distillation-a-comprehensive-study-
Deep Neural Networks
Mentioned in 2 entries
- cross-paradigm-knowledge-distillation-a-comprehensive-study-
- modeling-of-asdtd-childrens-behaviors-in-interaction-with-a-
FLUX.1-schnell
Mentioned in 2 entries
- chroma1-hd-character-transfer-with-flux2-dev
- they-want-to-rival-midjourney-so-here-you-go-chroma-v48-and-
NoobAI
Mentioned in 2 entries
- illustrious-noobai-style-explorer-5000-danbooru-artist-style
- illustriousnoob-ai-danbooru-tagging-getting-split-up
StableDiffusion
Mentioned in 2 entries
Google Gemma
Mentioned in 2 entries
- 4-mins-plus-for-a-response-to-simple-reply-with-ok
- googlegemma-in-space-surprise-guest-from-starcloud-and-aweso
Forge Neo
Mentioned in 2 entries
Zeta-Chroma
Mentioned in 2 entries
- getting-hilariously-bad-results-with-zeta-chroma-and-ernie-b
- is-there-a-complication-with-zeta-chromas-training
SD 1.5
Mentioned in 2 entries
- paperdoll-local-first-character-customization-for-vnindie-de
- watale-a-free-fully-local-visual-novel-engine-powered-by-sd-
K2.6
Mentioned in 2 entries
- been-working-with-kimimoonshot-k26-for-the-past-2-hours-acci
- ive-been-a-k25-superfan-since-it-came-out-these-new-numbers-
DeepSeekv4
Mentioned in 2 entries
GB300
Mentioned in 2 entries
Mosaic
Mentioned in 2 entries
- sparse-attention-to-the-details-preserving-spectral-fidelity
- today-we-release-mosaic-a-probabilistic-weather-model-that-s
variational autoencoder
Mentioned in 2 entries
- discovering-and-decoding-latent-mean-field-structure-with-va
- drivetrain-simulation-using-variational-autoencoders
flow matching
Mentioned in 2 entries
GPT-X
Mentioned in 2 entries
- agi-is-gpt-x-doing-all-of-the-work-after-you-say-we-would-lo
- asi-is-gpt-x-deciding-i-want-to-throw-a-party-for-myself-and
talkie-1930-13b-base
Mentioned in 2 entries
- introducing-talkie-a-13b-vintage-language-model-from-1930
- yeah-my-openclaw-runs-entirely-on-talkie-1930-13b-base-i-con
Z-Anime
Mentioned in 2 entries
Tuna-2
Mentioned in 2 entries
- meta-is-about-to-release-a-pixel-space-model-tuna-2
- tuna-2-pixel-embeddings-beat-vision-encoders-for-multimodal-
HappyHorse 1.0
Mentioned in 2 entries
- comfyui-now-supports-happyhorse-10-alibabas-cinematic-video-
- happyhorse-10-four-shot-anime-sequence-with-character-consis
Adonis
Mentioned in 2 entries
- adonis---general-consistencyupscale-edit-model-for-flux-2-kl
- update-adonis-post-model---for-adonis---general-consistencyu
Mimo V2.5 Pro
Mentioned in 2 entries
- bro-wtf-qwen-lab-cooked-with-qwen-38-27b-its-so-fucking-good
- quando-teremos-um-modelo-mimo-v25-pro-disponvel-na-cloud
Kimi k2.6
Mentioned in 2 entries
Qwen 3 14B
Mentioned in 2 entries
- no-s-nada-de-ollama-ni-programacin-ni-idea-pero-estoy-creand
- rtx-5080-with-16-gb-vram-64-gb-ram-best-quantized-model-for-
NoobAI-XL
Mentioned in 2 entries
- illustrious-noobai-style-explorer-now-with-16000-danbooru-ar
- looking-for-deleted-coco-style-noobai-xl--v60-checkpoint
SCAIL
Mentioned in 2 entries
- 7-seconds-made-with-7-ai-models-in-comfyui-after-effects-wan
- wan-animate-vs-wan-scail-scail-which-do-you-prefer-side-by-s
llama3.2:3b
Mentioned in 2 entries
- i-benchmarked-the-4-models-i-had-pulled-the-11gb-one-beat-th
- ollama-giving-outputs-as-json-not-working-right
Stable Diffusion 3.5
Mentioned in 2 entries
Sulphur 2
Mentioned in 2 entries
Wan
Mentioned in 2 entries
RTX 5090
Mentioned in 2 entries
NeMo
Mentioned in 2 entries
- 247-simulation-loops-how-agentic-ai-keeps-subsurface-enginee
- single-shot-generation-still-surfaces-net-new-kernels-with-n
nGPT
Mentioned in 2 entries
R1
Mentioned in 2 entries
- have-been-using-this-model-all-day-yesterday-with-its-new-ar
- when-to-retrieve-during-reasoning-pay-attention-to-this-one-
S4D
Mentioned in 2 entries
- beyond-diagonal-state-space-models-exact-non-abelian-group-t
- safety-security-and-cognitive-risks-in-state-space-models-a-
MedCLIP
Mentioned in 2 entries
- look-what-the-probes-dragged-in-real-world-chest-x-ray-short
- medsae-dissecting-medclip-representations-with-sparse-autoen
Seedance 2
Mentioned in 2 entries
SAM 3D
Mentioned in 2 entries
TRELLIS
Mentioned in 2 entries
- it-appears-that-microsoft-uploaded-an-image-model-on-hugging
- stream3d-sequential-multi-view-3d-generation-via-evidential-
AniMatrix
Mentioned in 2 entries
- animatrix-an-anime-video-generation-model-that-thinks-in-art
- tencent-is-about-to-release-an-anime-video-model-animatrix
Gradient Boosting
Mentioned in 2 entries
- conformal-risk-prediction-for-non-alcoholic-fatty-liver-dise
- sequential-structure-in-intraday-futures-data-lstm-vs-gradie
gpt-realtime-2
Mentioned in 2 entries
- fable-5-is-an-absolute-beast-at-threejs-its-the-best-llm-at-
- gpt-realtime-2-is-a-great-voice-model-with-a-typically-bad-o
GBDT
Mentioned in 2 entries
- balancing-trial-and-reorder-a-hybrid-sequential-transformer-
- nsr-boost-a-neuro-symbolic-residual-boosting-framework-for-i
Claude Opus 4.5
Mentioned in 2 entries
- completely-agreed-cost-management-was-not-a-priority-prior-t
- with-release-of-deepseek-v4-i-wanted-see-how-the-model-sizes
Rime Mist v3
Mentioned in 2 entries
- introducing-rime-mist-v3-on-together-ai-a-production-tts-fam
- try-rime-mist-v3-in-voice-finder-directly-httpsfindtherightv
Antigravity
Mentioned in 2 entries
- google-upgrades-notebooklm-which-now-runs-on-gemini-35-and-a
- the-crazy-thing-is-that-antigravity-does-this-quite-well-so-
Chatterbox TTS
Mentioned in 2 entries
Veo 3.1
Mentioned in 2 entries
- culturescore-evaluating-cultural-faithfulness-in-video-gener
- worldreasonbench-human-aligned-stress-testing-of-video-gener
Latent Consistency Models
Mentioned in 2 entries
FLUX.1
Mentioned in 2 entries
Carbon-3B
Mentioned in 2 entries
JoyAI-Echo
Mentioned in 2 entries
FLUX.1-dev
Mentioned in 2 entries
- i-combined-flux-fill-with-controlnet-for-structured-inpainti
- training-free-occluded-text-rendering-via-glyph-priors-and-a
AsymFLUX.2-klein-9B
Mentioned in 2 entries
FLUX.2-klein-base-9B
Mentioned in 2 entries
- asymflux2-klein-9b-is-all-about-textures
- flux2-klein-base-9b-controllight-lora-release-for-changing-l
GraphPFN
Mentioned in 2 entries
- graphpfn-a-prior-data-fitted-graph-foundation-model
- logic-budgeted-context-construction-for-node-level-graph-in-
Visual Geometry Transformer
Mentioned in 2 entries
- pacevggt-pre-alternating-attention-token-pruning-for-visual-
- visual-geometry-transformer-in-the-wild-distractor-free-3d-r
SCAIL-2
Mentioned in 2 entries
Granite
Mentioned in 2 entries
Lens
Mentioned in 2 entries
TinyLlama
Mentioned in 2 entries
- i-benchmarked-tinyllama-and-llama323b-on-boolean-logic-both-
- junk-responses-out-of-larger-models-but-tinyllama-seems-fine
YOLO-World
Mentioned in 2 entries
- vlod-tta-test-time-adaptation-of-vision-language-object-dete
- you-only-landmark-once-lightweight-u-net-face-super-resoluti
WAN 2.1
Mentioned in 2 entries
- can-you-use-a-normal-14b-lora-with-the-wan-14b-aio-model
- i-compared-62-samplers-and-16-schedulers-for-wan-21-image-ge
Lance
Mentioned in 2 entries
- lance-by-bytedance-3b-apache2-model-for-image-and-video-unde
- lance-unified-multimodal-modeling-by-multi-task-synergy
LFM2.5-8B-A1B
Mentioned in 2 entries
Visual Geometry Grounded Transformer
Mentioned in 2 entries
- not-all-tasks-quantize-equally-fisher-guided-quantization-fo
- qvggt-post-training-quantized-visual-geometry-grounded-trans
SWE-check
Mentioned in 2 entries
- today-were-releasing-swe-check-a-specialized-bug-detection-m
- use-quick-review-for-a-fast-local-check-before-pushing-free-
Marlin
Mentioned in 2 entries
- ai-is-watching-a-film-via-marlinvisuals-whisperaudio-and-pal
- product-designer-httpssakanaaicareersproduct-designer-sakana
HiDream O1
Mentioned in 2 entries
Krea 2 Medium
Mentioned in 2 entries
Ideogram v4
Mentioned in 2 entries
Ace-Step 1.5
Mentioned in 2 entries
MaskGIT
Mentioned in 2 entries
- demystifying-maskgit-sampler-and-beyond-adaptive-order-selec
- masked-generative-transformer-is-what-you-need-for-image-edi
Granite Embedding Multilingual R2
Mentioned in 2 entries
IP-Adapter
Mentioned in 2 entries
Gemma 4 MoE
Mentioned in 2 entries
- free-the-100b-gemma-4-moe-gemini-flash-35-is-out-so-now-you-
- v0311-mlx-tighten-up-gemma4-moe-loading-code-16964
gpt-oss:120b
Mentioned in 2 entries
- codexs-computer-use-not-working-with-ollama
- qwen36-vs-gpt-oss120b-on-apple-silicon-three-qwen-variants-b
Anima 1.0
Mentioned in 2 entries
- any-anima-10-fine-tunes-with-a-quality-bump-while-being-styl
- official-turbo-lora-for-anima-10-has-been-posted
Alibaba Wan 2.2
Mentioned in 2 entries
- human-in-the-loop-made-in-comfyui-with-a-laundry-list-of-too
- locked-n-loaded-made-in-comfyui-adobeae-photoshop-suno-vid-m
qwen2.5-coder
Mentioned in 2 entries
- big-update-my-local-flask-coding-assistant-now-has-5-dev-mod
- i-benchmarked-tool-call-reliability-on-qwen25-coder-against-
Spiking Neural Network
Mentioned in 2 entries
- on-the-evaluation-of-spiking-neural-network-configurations-f
- online-continual-learning-on-intel-loihi-2-via-a-co-designed
Juggernaut Z
Mentioned in 2 entries
Wan 2.2 Remix
Mentioned in 2 entries
- an-almost-complete-lack-of-motion-in-wan-22-remix-generation
- wan-22-remix-is-the-best-for-uncensored-video-or-is-there-so
Qwen3.5 35B
Mentioned in 2 entries
- extened-garlic-to-run-qwen35-35b-a3b-float8-at-55-toks-on-rt
- qwen35-35b-a3b-uncensored-heretic-native-mtp-preserved-is-ou
Gemma 4:12b
Mentioned in 2 entries
Luna xHigh
Mentioned in 2 entries
- i-really-want-to-get-artisanal-with-my-model-selection-luna-
- the-labs-should-launch-some-easy-to-view-comparisons-of-thei
MOSS-Transcribe-Diarize-0.9B
Mentioned in 2 entries
- moss-transcribe-diarize-09b-is-now-open-source-on-huggingfac
- moss-transcribe-series-updateopenmoss-alongside-moss-transcr
VibeVoice-ASR-BitNet
Mentioned in 2 entries
VibeVoice-ASR
Mentioned in 2 entries
TRELLIS.2
Mentioned in 2 entries
- kaininja-extending-native-3d-generators-to-the-part-level
- trellis2-can-now-generate-a-high-quality-3d-asset-in-under-7
qwen3.5
Mentioned in 2 entries
Cosmos3-Super-Image2Video
Mentioned in 2 entries
- cosmos3-super-image2video-running-locally-on-a-single-rtx-pr
- nvidia-releases-cosmos3-super-image2video-64b-parametres
FLUX.2 klein
Mentioned in 2 entries
Qwen 3.6 35B A3B
Mentioned in 2 entries
- how-long-can-i-expect-to-wait-until-the-local-30b-a3b-fronti
- qwen3635b-a3b-coding-nvfp4-is-not-getting-image-data-from-ol
Crys-JEPA
Mentioned in 2 entries
- how-do-we-design-materials-with-ai-excited-to-introduce-crys
- i-will-present-tomorrow-our-recent-work-on-crys-jepa-where-w
MedNeXt
Mentioned in 2 entries
- comnet-a-mednext-corrdiff-framework-for-multi-site-brain-tum
- topology-driven-fusion-of-nnu-net-and-mednext-for-accurate-b
SAM-Audio
Mentioned in 2 entries
- listen-look-and-learn-learning-without-forgetting-through-sa
- when-audio-separation-hurts-zero-shot-asr-evaluating-sam-aud
GridSFM
Mentioned in 2 entries
- gridsfm-a-foundation-model-for-solving-ac-optimal-power-flow
- gridsfm-a-new-small-foundation-model-for-the-electric-grid
MJEPA
Mentioned in 2 entries
- mjepa-a-simple-and-scalable-joint-embedding-predictive-archi
- our-paper-mjepa-has-been-accepted-at-eccv2026-huge-thanks-to
I-JEPA
Mentioned in 2 entries
- applying-jepa-style-predictive-learning-to-ja4-derived-netwo
- text-conditional-jepa-for-learning-semantically-rich-visual-
Minimax Music 3
Mentioned in 2 entries
- minimax-music-3-open-weight-release-soon
- today-i-have-unsubscribed-from-suno-thanks-to-minimax-music
Tenet
Mentioned in 2 entries
- congrats-to-harvey-on-tenet-their-first-model-and-a-mileston
- introducing-tenet-our-first-model-post-trained-for-legal-ten
Vision Foundation Models
Mentioned in 2 entries
- industrial-synthetic-segment-pre-training
- the-perception-physics-paradox-probing-scientific-alignment-
LFM2.5-VL-3B
Mentioned in 2 entries
Conditional GANs
Mentioned in 2 entries
- em-gansim-real-time-and-accurate-em-simulation-using-conditi
- targeted-label-flipping-and-oversampling-attacks-on-federate
Laguna S 2.1
Mentioned in 2 entries
- ill-say-it-plainly-hermes-agent-desktop-is-the-best-agentic-
- weve-gotten-some-great-medium-sized-models-lately-dsv4-flash
Cycle-Consistent Adversarial Networks
Mentioned in 2 entries
- unsupervised-metal-artifact-reduction-in-dental-cbct-using-f
- where-the-cost-falls-a-deployment-aware-adoption-order-for-s
Mobius-v0
Mentioned in 2 entries
- intern-s2-mobius-foundation-model-with-decoupled-knowledge-a
- paper-intern-s2-mobius-foundation-model-with-decoupled-knowl
Ornith-1.0
Mentioned in 2 entries
- ornith-10-self-scaffolding-llms-for-agentic-coding
- ornith-397b-running-at-q4-on-a-single-rtx-pro-6000-blackwell
gemma4
Mentioned in 2 entries
3D CNN
Mentioned in 2 entries
- a-multi-dataset-benchmark-of-multiple-instance-learning-for-
- classification-of-covid-19-cases-from-chest-ct-volumes-using
Cosmos 3 Edge
Mentioned in 2 entries
- as-always-cosmos-3-edge-is-fully-open-this-includes-model-we
- post-train-nvidia-cosmos-3-edge-for-on-device-robot-control
Qwen3-VL-32B
Mentioned in 2 entries
- it-looks-like-minimax-h3-uses-qwen3-vl-32b-as-text-encoder-a
- spongebob-minimax-h3-test-4-x-5-seconds-turbo-6-steps-134476
Qwen3-VL 8B
Mentioned in 2 entries
- minimax-h3-with-a-4b-or-8b-text-encoder-instead-of-the-32b-u
- qwen3-vl-8b-on-a-laptop-vs-opus-55-sonnet-5-gpt-56-on-137-me
TielCoder
Mentioned in 2 entries
- cybertiel-35b-a3bs-uncensored-4-bit-quant-beats-opus-46-medi
- tielcoders-22-gb-4-bit-quant-matches-opus46-medium-on-recent
LLaVA-1.5-7B
Mentioned in 2 entries
- encoded-but-disconnected-decomposing-vision-language-model-f
- targeting-the-attention-heads-behind-object-hallucination-in
transformer
Mentioned in 2 entries
- circuit-synchronization-precedes-generalization-a-causal-pre
- human-like-fleeting-memory-improves-language-learning-but-im
NEXUS
Mentioned in 2 entries
- a-lightweight-foundation-model-for-collider-physics-with-mul
- fundamentals-large-tabular-model-nexus-is-now-available-on-a
Vision-Language Models (VLMs)
Mentioned in 2 entries
- dial-decoupling-intent-and-action-via-latent-world-modeling-
- guidedvla-specifying-task-relevant-factors-via-plug-and-play
RWKV-7
Mentioned in 2 entries
- gated-deltanet-2-is-almost-exactly-rwkv-7s-dplr-recurrence-n
- pararnn-parallel-training-for-non-linear-rnns-slstm-rwkv-7-c
ResNet18
Mentioned in 2 entries
- per-pixel-bounding-box-regression-dbscan-for-handwritten-wor
- real-time-pedestrian-attribute-recognition-with-yolov8-and-r
Diffusion Large Language Models
Mentioned in 2 entries
- focus-dllms-know-how-to-tame-their-compute-bound
- roll-out-and-roll-back-diffusion-llms-are-their-own-efficien
SEAL
Mentioned in 2 entries
- a-hybrid-tgn-seal-model-for-dynamic-graph-link-prediction
- bokbo-best-of-k-bad-options-calibrated-abstention-for-vla-po
GPT Images 2.0
Mentioned in 2 entries
Denoising Diffusion Probabilistic Models
Mentioned in 2 entries
- domain-adaptive-zero-shot-image-enhancement-via-locality-con
- unsupervised-anatomical-feature-learning-via-diffusion-model
OlmoEarth v1.1
Mentioned in 2 entries
- olmoearth-v11-a-more-efficient-family-of-models
- olmoearth-v11-a-more-efficient-family-of-olmoearth-models
Wan2.2-I2V
Mentioned in 2 entries
- built-an-open-source-one-prompt-to-cinematic-reel-pipeline-o
- timestep-aware-svdquant-gptq-for-w4a4-quantization-of-wan22-
NVIDIA Vera Rubin NVL72
Mentioned in 2 entries
- ai-compute-is-going-to-orbit-spacexs-starmind-ai1-satellite-
- nvidia-ceo-jensen-huang-at-dell-technologies-world-demand-is
Khala
Mentioned in 2 entries
- khala-scaling-acoustic-token-language-models-toward-high-fid
- you-can-now-make-mac-generate-high-quality-songs---ported-kh
Qwen 3.x
Mentioned in 2 entries
NVIDIA AVO
Mentioned in 2 entries
- nvidia-avo-reaches-100-on-arc-agi-3-demonstrating-a-frontier
- our-general-purpose-coding-agent-just-scored-100-on-the-arc-
M5 Pro
Mentioned in 2 entries
NVIDIA Alpamayo 2 Super
Mentioned in 2 entries
- generate-trajectories-reasoning-traces-and-auto-labels-with-
- nvidia-alpamayo-2-super-the-frontier-open-model-for-robotaxi
Blackwell
Mentioned in 2 entries
- eli-lilly-plans-to-use-its-73b-cash-reserve-to-fund-an-app-s
- the-most-important-question-hardly-anyone-is-asking-is-wheth
AraBERT
Mentioned in 2 entries
- arassm-a-bidirectional-state-space-encoder-for-arabic-masked
- extractive-summarization-for-arabic-documents-using-sarabert
Groq 3 LPX
Mentioned in 2 entries
- how-nvidia-groq-3-lpx-deterministic-execution-drives-power-e
- how-nvidia-groq-3-lpx-unlocks-ultrafast-interactivity-at-lon
GPT-2 Small
Mentioned in 2 entries
- reading-task-failure-off-the-activations-a-sparse-feature-au
- the-cognitive-categorical-transformer-category-theoretic-ind
RTX Spark
Mentioned in 2 entries
- microsofts-surface-laptop-ultra-looks-like-its-first-true-ma
- this-could-be-windows-m1-moment-but-expect-it-to-cost-a-ton
AlphaStar
Mentioned in 2 entries
- exclusive-david-silver-is-assembling-an-absurd-reinforcement
- runtime-action-interference-for-ai-control-of-alphastar-in-s
Masked Diffusion Models
Mentioned in 2 entries
- bringing-stability-to-diffusion-decomposing-and-reducing-var
- the-confidence-shortcut-a-reasoning-failure-mode-of-masked-d
Decision Transformers
Mentioned in 2 entries
- generalizable-multi-task-learning-for-wireless-networks-usin
- guided-riemannian-optimization-guro-bridging-model-predictiv
GPT-4o mini
Mentioned in 2 entries
- in-other-news-we-updated-the-55-instant-model-used-in-chatgp
- is-gpt-4o-mini-blinded-by-its-own-safety-filters-exposing-th
Wall-OSS-0.5
Mentioned in 2 entries
Pythia
Mentioned in 2 entries
- a-negative-result-on-cross-model-activation-transfer-in-a-py
- lagged-coupling-internal-representations-become-readable-bef
Qwen2.5-7B-Instruct
Mentioned in 2 entries
- peak-then-collapse-and-the-four-interface-channels-of-knowle
- the-shape-of-wisdom-decision-trajectories-in-language-models
DQN
Mentioned in 2 entries
- ensemble-elastic-dqn-a-step-dependent-ensemble-approach-for-
- mitigating-the-stability-plasticity-dilemma-in-adaptive-trai
Vision-Language-Action Models
Mentioned in 2 entries
- quovla-quotient-space-for-vision-language-action-models
- trust-your-instincts-confidence-driven-test-time-rl-for-visi
Nanbeige4.2-3B
Mentioned in 2 entries
- nanbeige42-3b-on-apple-silicon-fixing-deployment-bugs-and-de
- nanbeige42-3b-unlocking-agentic-capabilities-in-a-compact-mo
D-JEPA
Mentioned in 2 entries
- beyond-representation-learning-a-systematic-study-of-joint-e
- d-jepa-a-decision-aligned-latent-world-model
transformers
Mentioned in 2 entries
- n-gram-like-language-models-predict-naturalistic-reading-tim
- normalizing-flows-are-capable-models-for-continuous-control
Proximal Policy Optimization
Mentioned in 2 entries
- bridging-the-gap-enabling-soft-actor-critic-for-high-perform
- satellite-trajectory-optimization-via-proximal-policy-optimi
GPT-4V
Mentioned in 2 entries
- can-frontier-ai-models-actually-read-a-painting-r
- dual-adversarial-fine-tuning-for-enhancing-robustness-of-lar
Lean 4
Mentioned in 2 entries
- expected-value-alignment-for-generative-reward-modeling-in-f
- topoalign-a-framework-for-aligning-code-to-math-via-topologi
MADDPG
Mentioned in 2 entries
- enhancing-the-maddpg-algorithm-for-multi-agent-learning-via-
- scalable-neighborhood-based-multi-agent-actor-critic
RETFound
Mentioned in 2 entries
- fundusegmenter-leveraging-the-retfound-foundation-model-for-
- uncertainty-aware-last-layer-adaptation-of-retfound-for-refe
dLLM
Mentioned in 2 entries
- elastic-dllm-position-preserving-context-compression-and-aug
- learning-from-the-self-future-on-policy-self-distillation-fo
TelecomGPT-R1
Mentioned in 2 entries
- telecomgpt-r1-a-unified-open-source-reasoner-for-the-telecom
- telecomgpt-r1-unified-post-training-for-reasoning-across-het
Segment Anything Models
Mentioned in 2 entries
- structsam-structure--and-spectrum-preserving-token-merging-f
- towards-any-quality-image-segmentation-via-generative-and-ad
Multimodal LLMs
Mentioned in 2 entries
- from-generator-to-embedder-harnessing-innate-abilities-of-mu
- me-2023-2024-2025-multimodal-llms-are-going-to-have-problems
KAT-Coder-V2.5
Mentioned in 2 entries
NEAT
Mentioned in 2 entries
- investigating-hyperparameter-optimization-and-transferabilit
- neat-neighborhood-guided-efficient-autoregressive-set-transf
VLAs
Mentioned in 2 entries
PlanGPT
Mentioned in 2 entries
- a-complementary-study-on-plangpt-evaluation-with-defined-per
- symmetry-aware-transformer-training-for-automated-planning
ZAYA1-8B
Mentioned in 2 entries
Linear Recurrent Units
Mentioned in 2 entries
- rotation-free-online-handwritten-character-recognition-using
- rotrnn-modelling-long-sequences-with-rotations
ESMC
Mentioned in 2 entries
- a-closer-look-at-what-we-released-today-esmc-is-a-language-m
- alexrives-on-ai-for-science-with-latentspacepod-breaking-dow
Diffusion Probabilistic Models
Mentioned in 2 entries
- high-fidelity-synthetic-transmission-electron-microscopy-ima
- synboost-a-synergistic-framework-for-fast-sampling-of-diffus
CBAM
Mentioned in 2 entries
- spatiality-frequency-domain-video-forgery-detection-system-b
- uncertainty-aware-wildfire-smoke-density-classification-from
Qwen2.5-VL-7B
Mentioned in 2 entries
- encoded-but-disconnected-decomposing-vision-language-model-f
- probe-choice-changes-canary-memorization-verdicts-three-post
Deep Belief Network
Mentioned in 2 entries
- automatic-extraction-of-road-networks-by-using-teacher-stude
- classifying-directional-trajectories-near-criticality-in-the
EAGLE-3
Mentioned in 2 entries
- margins-not-windows-training-free-per-step-lossy-speculative
- speculative-decoding-implementations-eagle-3-medusa-1-pard-d
LeVLJEPA
Mentioned in 2 entries
- oops-sigreg-did-it-again-large-scale-cc12m-datacomp-l-vision
- we-introduce-levljepa-the-first-fully-non-contrastive-end-to
Neural Ordinary Differential Equations
Mentioned in 2 entries
- frequency-domain-neural-odes-for-modeling-non-linear-dynamic
- virtual-temperature-sensors-in-power-transformers-using-neur
Minimax M2.7
Mentioned in 2 entries
- budget-blown-on-closed-apis-is-a-solvable-problem-simply-rep
- here-is-a-2nd-batch-of-april-architecture-drops-what-a-month
consistency models
Mentioned in 2 entries
- flow-map-grpo-reinforcement-learning-for-few-step-flow-map-g
- flowmap-opd-rollout--kernel-separation-for-on-policy-distill
Probabilistic Transformer
Mentioned in 2 entries
- exploring-the-potential-of-probabilistic-transformer-for-tim
- scaling-probabilistic-transformer-via-efficient-cross-scale-
GPTQ
Mentioned in 2 entries
- gptq-2d-cubic-time-two-sided-adaptive-rounding
- paper-statistically-lossless-quantization-of-large-language-
Liquid Neural Network
Mentioned in 2 entries
- liquid-neural-network-models-for-natural-gas-spot-price-time
- multi-rate-mixture-of-experts-for-accelerating-liquid-neural
QLSTM
Mentioned in 2 entries
- recursive-qlstm-with-dynamic-variational-quantum-circuit-ada
- recursive-quantum-long-short-term-memory-for-stable-short-ho
LEMUR
Mentioned in 2 entries
- lemur-latent-entropy-aware-multimodal-unlearning-via-visual-
- systematic-exploration-of-4-expert-heterogeneous-mixture-of-
Aleph
Mentioned in 2 entries
- ebms-are-so-back-aleph-now-leads-the-major-formal-reasoning-
- we-built-aleph-as-one-of-the-first-generalist-text-based-edi
LLaDA2.0-Uni
Mentioned in 2 entries
AlphaGeometry
Mentioned in 2 entries
- i-want-ai-to-succeed-and-have-been-positive-about-many-syste
- the-topological-dual-of-a-dataset-a-logic-to-topology-encodi
wav2vec 2.0
Mentioned in 2 entries
- layer-wise-probing-of-wav2vec-20-and-whisper-for-consonant-c
- the-ntnu-system-at-the-si-challenge-2025-sla-open-track
CNN-BiLSTM
Mentioned in 2 entries
- a-comparative-study-of-pycaret-automl-and-cnn-bilstm-for-bin
- sedentary-behavior-classification-for-wearable-sensors-with-
3D Convolutional Neural Networks
Mentioned in 2 entries
- deepfake-detection-in-social-media-a-temporal-artifact-analy
- online-hand-gesture-recognition-using-3d-convolutional-neura
Generative Adversarial Networks
Mentioned in 2 entries
- brain-to-image-generation-reconstructing-visual-stimuli-from
- deep-learning-based-on-generative-adversarial-and-convolutio
3D Gaussian Splat
Mentioned in 2 entries
- deblurnvs-geometric-latent-diffusion-for-novel-view-synthesi
- fast-and-lightweight-novel-view-synthesis-with-differentiabl
video diffusion models
Mentioned in 2 entries
- mirrorworld-taming-video-diffusion-models-for-mirror-reflect
- stabilizing-camera-controlled-novel-view-synthesis-at-infere
SRGAN
Mentioned in 2 entries
- enhancing-ebsd-throughput-of-battery-electrode-materials-usi
- srgan-ckan-expressive-super-resolution-with-nonlinear-functi
IndoBERT
Mentioned in 2 entries
- benchmarking-logistic-regression-svm-naive-bayes-and-indober
- benchmarking-pycaret-automl-against-indobert-fine-tuning-for
XLM-RoBERTa
Mentioned in 2 entries
- biosentinel-at-exist-2026-soft-label-optimization-with-xlm-r
- ncl-bu-at-semeval-2026-task-3-fine-tuning-xlm-roberta-for-mu
ES-HyperNEAT
Mentioned in 2 entries
- early-stopping-thresholds-for-es-hyperneat-a-data-driven-app
- investigating-hyperparameter-optimization-and-transferabilit
Perceiver
Mentioned in 2 entries
- courant-a-state-adaptive-perceiver-based-neural-surrogate-wi
- hipervit-a-hierarchical-perceiver-vision-transformer-archite
Sakana Fugu Ultra
Mentioned in 2 entries
- sakana-fugu-ultra-is-live-on-ai-gateway-mythos-class-intelli
- sakana-fugu-ultra-now-available-on-ai-gateway
Bidirectional LSTM
Mentioned in 2 entries
- manipulation-of-deformable-linear-objects-using-model-predic
- state-of-health-estimation-using-convolutional-and-bidirecti
FunSearch
Mentioned in 2 entries
- effective-harness-engineering-for-algorithm-discovery-with-c
- llm-guided-search-for-deletion-correcting-codes
wav2vec2
Mentioned in 2 entries
- speaker-labeled-transcription-with-whisperx-on-sagemaker-ai
- wav2vot-automatic-estimation-of-voice-onset-time-closure-dur
VGG19
Mentioned in 2 entries
- a-comparative-study-of-gan-based-deep-learning-models-for-pn
- classification-of-disease-from-lungs-x-ray-images-using-vgg1
WGAN-GP
Mentioned in 2 entries
- cognitive-energy-modeling-for-neuroadaptive-human-machine-sy
- gan-diff-coupling-pretrained-wgan-gp-features-with-condition
Dense Associative Memory
Mentioned in 2 entries
- algorithmic-analysis-of-dense-associative-memory-finite-size
- geometric-entropy-and-retrieval-phase-transitions-in-continu
NeRV
Mentioned in 2 entries
- lrconv-nerv-low-rank-convolution-for-efficient-neural-video-
- rt-nerv-rethinking-hybrid-neural-representations-for-video-v
Temporal Fusion Transformer
Mentioned in 2 entries
- deep-learning-surrogates-for-emulating-stochastic-climate-ti
- spatially-enhanced-temporal-fusion-transformer-interpretable
Boltz-2
Mentioned in 2 entries
- adapting-boltz-2-with-limited-experimental-activity-data-imp
- modeling-atomic-conformational-ensembles-of-proteins-via-tes
3DMM
Mentioned in 2 entries
- a-large-scale-evaluation-of-text-guided-models-for-facial-ed
- fields-face-reconstruction-with-accurate-inference-of-expres
DCGAN
Mentioned in 2 entries
- dcgan-inference-on-a-microcontroller-126m-parameters-512kb-s
- does-dcgan-based-synthetic-augmentation-improve-brain-tumor-
ImageNet
Mentioned in 2 entries
- rethinking-transfer-learning-for-industrial-inspection-dinov
- training-a-predictive-coding-network-on-imagenet-using-equil
Liquid Neural Networks
Mentioned in 2 entries
- liquid-neural-networks-as-a-drop-in-continuous-time-deformat
- mtliquid-enabling-efficient-multi-task-learning-using-liquid
MetaFormer
Mentioned in 2 entries
- mradnet-a-compact-radar-object-detector-with-metaformer
- shaken-or-stirred-an-analysis-of-metaformers-token-mixing-fo
TabPFN-TS
Mentioned in 2 entries
- investigating-simple-target-covariate-relationships-for-chro
- systematic-evaluation-of-tabpfn-ts-for-zero-shot-probabilist
Jamba
Mentioned in 2 entries
- ai21-achieves-an-83-reduction-in-time-to-start-for-ai-worklo
- asymmetric-virtual-memory-paging-for-hybrid-mamba-transforme
Extreme Learning Machine
Mentioned in 2 entries
- modelling-non-linear-aeroelastic-loads-in-long-span-bridges-
- spectral-stability-of-pseudoinverse-based-extreme-learning-m
ESMFold
Mentioned in 2 entries
- eladgil-bs-attention-was-born-in-montral-pytorch-in-nyc-alph
- two-stages-of-folding-convergent-mechanisms-in-ai-protein-fo
Dreamer
Mentioned in 2 entries
- dreamer-cdp-improving-reconstruction-free-world-models-via-c
- one-lens-many-worlds-a-capability-typed-interface-for-world-
IndicTrans2
Mentioned in 2 entries
- conversational-domain-adaptation-of-indictrans2-across-21-in
- english-to-prakrit-machine-translation-via-multilingual-tran
Qwen3 0.6B
Mentioned in 2 entries
- react-native-executorch-is-now-up-to-92x-faster
- where-are-small-models-like-qwen3-06b-and-qwen35-08b-used-hu
World Action Models
Mentioned in 2 entries
- learning-to-use-imagination-progress-conditioned-future-util
- q-wam-4-bit-quantization-of-world-action-models-with-action-
Latent diffusion models
Mentioned in 2 entries
- keep-or-drop-adaptive-tokenizer-for-compact-video-representa
- what-makes-a-good-medical-image-tokenizer-rethinking-reconst
Anthropic Opus 4.8
Mentioned in 2 entries
- anthropic-opus-48-is-new-sota-on-arc-agi-3-score-15-10k-arc-
- deepseek-unveils-an-experimental-multimodal-version-of-its-v
agnost-0.1
Mentioned in 2 entries
MiniMax-M2
Mentioned in 2 entries
- run-it-on-the-ai-native-cloud-serverless-and-dedicated-infra
- the-minimax-m2-series-mini-activations-unleashing-max-real-w
Qwen3.6-Max-Preview
Mentioned in 2 entries
- introducing-qwen36-max-preview-an-early-preview-of-our-next-
- qwen37-max-just-landed-at-566-on-the-artificial-analysis-int
ERNIE 5.1
Mentioned in 2 entries
- ernie-51-is-here-ernie-51-significantly-reduces-pretraining-
- i-like-that-there-are-models-called-bert-and-ernie-but-in-al
Devin Fusion
Mentioned in 2 entries
- conventional-model-routing-sucks-it-passes-benchmarks-but-fa
- watch-devin-fusion-work-the-sidekick-model-handles-fetching-
Trace Judge
Mentioned in 2 entries
- research-product-very-excited-to-start-rolling-out-our-fine-
- we-are-starting-to-role-out-our-trace-judge-model-to-early-p
hermes-agent
Mentioned in 2 entries
GPT 5.4
Mentioned in 2 entries
- dear-god-lol---new-kimi-model-is-a-fucking-beast-gpt-54-leve
- kimi-k26-just-entered-the-chat-beats-gpt-54-and-opus-46-on-c
North Mini Code
Mentioned in 2 entries
- introducing-coheres-first-open-source-coding-model-north-min
- introducing-north-mini-code-coheres-first-model-for-develope
Masked Diffusion Language Models
Mentioned in 2 entries
- masked-diffusion-language-models-are-strong-and-steerable-te
- membership-inference-attacks-on-discrete-diffusion-language-
SWE-Pro
Mentioned in 2 entries
- on-software-engineering-5622-swe-pro-matching-gpt-53-codex-5
- sakana-aiaisakana-fugu-httpssakanaaifugu-betajapanese
SWE 1.6 Fast
Mentioned in 2 entries
- missed-the-first-100-httpsxcommattschragestatus2052502157396
- swe-16-fast-is-available-in-devin-for-terminal-httpdevinaite
Qwen3.8-Max-Preview
Mentioned in 2 entries
- during-preview-qwen38-is-getting-better-by-the-day-latest-ve
- since-launching-qwen38-max-preview-weve-received-valuable-fe
Muse Code
Mentioned in 2 entries
- introducing-muse-code-and-muse-spark-12
- meta-takes-on-anthropic-and-openai-with-its-first-ai-coding-
Trinity-Large-Thinking
Mentioned in 2 entries
- one-of-the-great-parts-of-nous-portal-is-stuff-like-this
- trinity-large-thinking-arceeais-latest-model-is-now-free-on-
Qwen3.5-35B-A3B
Mentioned in 2 entries
- lm-performanceqwen36-35b-a3b-outperforms-the-dense-27b-param
- we-built-a-local-ai-work-tool-that-runs-qwen35-35b-a3b-on-a-
Nano-Banana
Mentioned in 2 entries
Qwen 3.5 35B
Mentioned in 2 entries
- hermes-nousresearch-agent-with-qwen3535b-a3b-on-a-4090-is-ve
- running-qwen-35-35b-a3b-q80-gguf-on-a-cheap-radeon-7600-at-1
Granite 4.2
Mentioned in 2 entries
Sa2VA
Mentioned in 2 entries
- agentrvos-for-mevis-text-track-of-5th-pvuw-challenge-3rd-met
- event-driven-refresh-and-recurrence-memory-to-reduce-stale-g
o3-mini
Mentioned in 2 entries
- banger-paper-from-stanford-and-together-ai-they-show-why-it-
- o3-mini-very-obsolete-in-an-agentic-loop-made-good-exam-ques
AgenticQwen-30B-A3B
Mentioned in 2 entries
- new-paper-from-alibaba-a-30b-moe-with-only-3b-active-params-
- the-top-ai-papers-of-the-week-april-26---may-3---latent-agen
GPT Image v2
Mentioned in 2 entries
- i-created-a-github-repo-with-top-gpt-image-v2-prompts-this-m
- whats-this-weird-texture-with-image-v2-why-does-it-make-the-
AlphaGenome
Mentioned in 2 entries
- google-deepmind-releases-alphagenome-atlas-a-1pb-dataset-of-
- google-deepminds-alphagenome-atlas-maps-all-9-billion-possib
Image 2
Mentioned in 2 entries
- i-asked-the-new-image-2-to-generate-fortnite-this-is-what-i-
- image-2-model-is-pretty-good-at-creating-promotional-product
5.6
Mentioned in 2 entries
- 56-livestream-going-now-in-addition-to-the-model-3-major-pro
- weve-usually-stayed-away-from-model-comparisons-but-56-vs-fa
FSD v14.3.2
Mentioned in 2 entries
- fsd-v1432-has-been-driving-me-all-over-la-tonight-this-build
- the-advantage-of-having-eyes-all-around
30B-A3B
Mentioned in 2 entries
- paper-of-the-day-httpshuggingfacecopapers260513301
- were-releasing-a-30b-a3b-reasoning-model-that-reaches-gold-m
Images 2.0
Mentioned in 2 entries
- images-20-really-got-over-some-important-qualitative-thresho
- the-new-image-generation-nails-text-much-more-accurately
AlphaFold2
Mentioned in 2 entries
- a-couple-of-months-ago-i-announced-that-i-was-partway-throug
- accelerator-choice-is-not-enough-alphafold2-inference-on-clo
Krea 2 Raw
Mentioned in 2 entries
- released-wulver-v05-our-krea-2-finetune-for-anime-and-furry
- today-we-release-the-open-weights-of-krea-2-welcome-krea-2-r
Autopilot
Mentioned in 2 entries
- elon-musk-denies-teslas-autopilot-caused-crash-that-killed-g
- fsd-is-shockingly-good-i-am-completely-gobsmacked-tesla-team
GPT-Live-1
Mentioned in 2 entries
- chatgpts-upgraded-voice-mode-is-better-at-shutting-up
- openai-rolls-out-two-versions-of-gpt-live-gpt-live-1-powerin
ChatGPT 2.0
Mentioned in 2 entries
Sora 2
Mentioned in 2 entries
Siri
Mentioned in 2 entries
- apple-reaches-a-250m-settlement-in-a-ca-federal-court-to-res
- tim-cook-and-eu-tech-chief-henna-virkkunen-held-constructive
Atlas
Mentioned in 2 entries
- fei-fei-lis-world-labs-debuts-atlas-a-world-model-showcase-f
- from-feed-forward-to-flow-unifying-reconstruction-and-genera
5.5
Mentioned in 2 entries
- there-are-early-signs-of-55-being-a-competent-ai-research-pa
- wow-yall-love-55-we-should-think-of-something-nice-to-do-to-
GPT Image 1.5
Mentioned in 2 entries
Seedance 2.5
Mentioned in 2 entries
- bytedance-unveils-seedance-25-in-beijing-saying-the-ai-video
- the-upcoming-seedance-25-model-looking-insane-as-well-with-m
Grok Heavy
Mentioned in 2 entries
- grok-heavy-analysis-httpsgrokcomsharebgvnywn537328302-c895-4
- holy-shit-im-going-to-need-another-grok-heavy-account-grok-b
Grok Imagine Video 1.5
Mentioned in 2 entries
Privacy Filter
Mentioned in 2 entries
- openai-just-open-sourced-a-new-15b-50m-active-model-on-huggi
- openai-releases-privacy-filter-an-open-weight-model-for-mask
GR00T
Mentioned in 2 entries
- sonic-training-code-finetuning-checkpoint-vla-data-collectio
- we-are-very-pleased-to-see-nvidiarobotics-asierarranz-using-
grok 4.5
Mentioned in 2 entries
- grok-45-is-seriously-good-its-going-in-the-direction-which-i
- grok-45-made-me-give-grok-build-a-serious-run-today-heres-my
Ideogram V4 instant
Mentioned in 2 entries
- weve-open-sourced-ideogram-v4-instant-and-fast-find-them-her
- weve-open-sourced-our-in-house-distilled-models-again-much-f
Ideogram V4 fast
Mentioned in 2 entries
- weve-open-sourced-ideogram-v4-instant-and-fast-find-them-her
- weve-open-sourced-our-in-house-distilled-models-again-much-f
GPT Pro
Mentioned in 2 entries
- number-1-openai-codex-request-plz-let-us-use-gpt-pro-in-it-o
- top-universities-in-south-korea-are-now-providing-students-w
FSD V14 Lite
Mentioned in 2 entries
- following-future-rollout-of-fsd-v14-lite-for-hw3-vehicles-in
- think-about-this-the-7-year-old-teslas-with-ai3-that-are-now
Grok Imagine Video
Mentioned in 2 entries
- 3-new-models-from-xais-grok-creative-stack-are-live-on-openr
- grok-imagine-video-has-become-the-clear-leader-in-vercel-ai-
GPT Image
Mentioned in 2 entries
- a-fix-is-being-worked-on-for-artifacting---boyuan-chen-core-
- invokeai-613-just-released-its-largest-community-driven-rele
LeRobot
Mentioned in 2 entries
- excited-to-share-that-i-joined-huggingface-as-a-ml-advocate-
- hf-is-the-perfect-place-to-host-your-big-robotics-datasets-r
Daybreak Red
Mentioned in 2 entries
- accelerate-cyber-defense-with-openai-and-aws-daybreak-red-da
- openai-unveils-two-daybreak-tiers-daybreak-blue-which-provid
GLM 5.2 Fast
Mentioned in 2 entries
- glm-52-fast-via-wafer-now-available-on-ai-gateway
- we-heard-your-feedback-you-want-to-go-faster-introducing-glm
TML-Interaction-Small 276B-A12B
Mentioned in 2 entries
- ainews-thinking-machines-native-interaction-models---tml-int
- my-highlights-httpsxcomlatentspacepodstatus20540578350431398
PP-OCRv6
Mentioned in 2 entries
- a-free-lunch-adapting-pp-ocrv6-for-historical-text-recogniti
- pp-ocrv6-on-hugging-face-50-language-ocr-from-15m-to-345m-pa
gpt-image-2
Mentioned in 2 entries
- true-to-form-ive-already-seen-openai-themselves-refer-to-the
- wheres-the-raccoon-with-the-ham-radio-chatgpt-images-20
Laguna S
Mentioned in 2 entries
- inside-the-model-factory-eiso-kant-poolside-ai
- other-active-promotions---free-models-solar-pro-4-1-week-hy3
Chatterbox
Mentioned in 2 entries
- praxy-voice-voice-prompt-recovery-bups-for-commercial-class-
- voice-cloning-is-style-transfer-across-three-widely-used-sys
Claude 3.5 Sonnet
Mentioned in 2 entries
- liteparse-runs-so-fast-that-claude-fable-5-doesnt-think-its-
- m3-brings-sparse-attention-1m-context-multimodality-and-toge
Qwen3.8-27B-BF16
Mentioned in 2 entries
Inflect v2
Mentioned in 2 entries
- i-released-inflect-v2-two-ultra-tiny-complete-tts-models-und
- you-can-now-fine-tune-my-396m-parameter-tts-on-your-own-voic
Support Vector Regression
Mentioned in 2 entries
- analytical-evaluation-of-dca-convergence-properties-for-mini
- optimised-support-vector-regression-for-california-housing-p
LISA
Mentioned in 2 entries
- medveriseg-teaching-mllm-based-medical-segmentation-models-t
- rose-retrieval-oriented-segmentation-enhancement
Qwen-Image-2.0
Mentioned in 2 entries
GPT-5 Pro
Mentioned in 2 entries
- im-very-excited-about-this-article-from-openai-on-my-attempt
- interesting-that-the-gpt-5-pro-series-models-have-consistent
North Micro Vision
Mentioned in 2 entries
- speaking-of-fine-tuning-weve-got-support-on-axolotlai-start-
- today-were-adding-another-member-to-our-model-family-meet-no
Ling 3.0 flash
Mentioned in 2 entries
- antling-open-sourced-ling-30-flash-fin-a-finance-enhanced-mo
- we-officially-surpassed-1m-model-downloads-on-huggingface-yo
ModelScope
Mentioned in 2 entries
DUSt3R
Mentioned in 2 entries
- interp3r-continuous-time-3d-geometry-estimation-with-frames-
- raytun3r-online-camera-adaptation-in-3d-foundation-models
Depth Anything 3
Mentioned in 2 entries
- dyna3-vlm-guided-training-free-4d-reconstruction-via-depth-f
- raytun3r-online-camera-adaptation-in-3d-foundation-models
Ternary Bonsai
Mentioned in 2 entries
- 1-bit-2-bit-ternary-bitnet-models---updates-tracking
- today-were-announcing-ternary-bonsai-top-intelligence-at-158
Ornith-1.5 9B
Mentioned in 2 entries
Gemini Robotics
Mentioned in 2 entries
- boston-dynamics-just-gave-its-robot-dog-a-brain-that-reasons
- google-deepminds-new-ai-model-can-control-a-robots-entire-bo
NuExtract3
Mentioned in 2 entries
- nuextract3-is-now-available-on-ollama-4b-vlm-for-document-to
- nuextract3-released-open-weight-4b-vlm-for-markdown-ocr-and-
Gemini Omni 1.1 Flash
Mentioned in 2 entries
- gemini-omni-11-flash-is-live-in-comfyui-via-partner-nodes-go
- meet-gemini-omni-11-flash-our-newest-multimodal-model-for-vi
PXDesign
Mentioned in 2 entries
RFdiffusion
Mentioned in 2 entries
LongCat-2.0
Mentioned in 2 entries
- chinas-meituan-open-sources-massive-longcat-20-ai-model-sayi
- meituan-open-sources-longcat-20-a-16t-parameter-model-that-i
glm-5.1
Mentioned in 2 entries
- ollama-launch-claude---model-glm-51cloud-we-are-rushing-to-g
- some-models-to-try-with-codex-kimi-k26cloud-with-vision-supp
Llama 3.2 1B
Mentioned in 2 entries
- fine-tuned-llama-32-1b-on-480-examples-shipped-to-android-vi
- klq-training-free-measured-rotation-quantization-beats-all-t
DeepSeek 4
Mentioned in 2 entries
DSV4
Mentioned in 2 entries
Qwen-Image
Mentioned in 2 entries
- position-paper-paired-ab-forgetting-on-purpose-five-tells-fo
- side-by-side-comparison-of-qwen-image-ernie-baseturbo-and-fl
Llama-3-8B
Mentioned in 2 entries
- option-order-randomisation-reveals-a-distributional-position
- quantisation-reshapes-the-metacognitive-geometry-of-language
GPT-5.3-Codex
Mentioned in 2 entries
- ai-reviewers-then-ranked-the-submissions-and-gave-the-same-o
- on-software-engineering-5622-swe-pro-matching-gpt-53-codex-5
Xiaomi MiMo-V2.5
Mentioned in 2 entries
- going-to-buy-2-rtx-6000-just-because-of-the-capabilities-of-
- xiaomi-mimo-v25-is-now-officially-open-sourced-mit-license-s
Qwen 3.6 35B MoE
Mentioned in 2 entries
- running-qwen-36-35b-moe-q4km-on-a-zeus-xiaomi-12-pro-12gb-ra
- running-qwen-36-35b-moe-with-zoo-code-on-m1-max-is-amazing-f
GLM5.3 Flash
Mentioned in 2 entries
TokaMind
Mentioned in 2 entries
- tokamind-a-multi-modal-transformer-foundation-model-for-toka
- tokamind-for-power-grid-cross-domain-transfer-from-fusion-pl
Inkling Small
Mentioned in 2 entries
- inkling-and-inkling-small-are-now-served-directly-by-thinkym
- weve-gotten-some-great-medium-sized-models-lately-dsv4-flash
ERNIE-Image-8b
Mentioned in 2 entries
- a-new-image-model-ernie-image-8b-from-baidu-will-be-released
- great-news-the-ernie-editing-model-is-expected-to-be-release
CLEF
Mentioned in 2 entries
Gemma 4 26b
Mentioned in 2 entries
Sol 5.6
Mentioned in 2 entries
Miso One
Mentioned in 2 entries
- another-banger-open-source-release-miso-one-is-an-8b-text-to
- today-were-excited-to-introduce-miso-one-the-most-emotive-vo
Claude Science
Mentioned in 2 entries
- claude-science-is-anthropics-newest-flagship-product
- the-download-anthropic-launches-claude-science-and-californi
Visual Mamba
Mentioned in 2 entries
- 3d-segment-anything-model-with-visual-mamba-for-diagnosing-p
- can-visual-mamba-improve-ai-generated-image-detection-an-in-
Gemini 2.5 Pro
Mentioned in 2 entries
DeepSeek-R1-Distill-Qwen-7B
Mentioned in 2 entries
- the-halt-vector-internalizing-a-causal-steering-intervention
- token-budget-saturation-and-mechanistic-early-detection-of-r
Kimi K2.7 Code
Mentioned in 2 entries
- the-underrated-part-of-this-announcement-is-that-fireworks-h
- to-celebrate-bionics-launch-were-offering-2-credits-for-a-li
Gemini 3.8 Flash Cyber
Mentioned in 2 entries
- google-launches-gemini-38-flash-cyber-for-partners-in-its-ne
- google-launches-two-gemini-38-models-with-cutting-edge-reaso
Gemini Flash 3.5
Mentioned in 2 entries
- free-the-100b-gemma-4-moe-gemini-flash-35-is-out-so-now-you-
- gemini-flash-35-is-such-a-disappointing-model-its-intelligen
Bart
Mentioned in 2 entries
Qwen3.8-Flash-Next-NVFP4
Mentioned in 2 entries
- local-agentic-coding-benchmark-qwen38-flash-next-nvfp4-vs-27
- qwen38-flash-next-nvfp4-vs-qwen38-27b-fp-test-results
GPT-5.6 S
Mentioned in 2 entries
- 10000-openai-agents-worked-88-hours-on-the-navier-stokes-cla
- grok-46-ranks-1-on-the-gpqa-diamond-leaderboard-grok-46-high
DeepSeek-V4 Flash 0731
Mentioned in 2 entries
- deepseek-v4-flash-0731-vs-gpt-56-luna-on-deepswe-cost-and-co
- glm-53-flash-aka-ox-alpha-is-now-the-best-intelligence-per-d
G9v3-39A5B
Mentioned in 2 entries
G9v3-3B
Mentioned in 2 entries
Qwen3-0.6B
Mentioned in 2 entries
- i-reverse-engineered-an-npu-vendors-engine-format-int8-weigh
- qwen3-06b-400-mb-on-a-samsung-note-8-2017-phone-drives-a-rea
TimesFM
Mentioned in 2 entries
- nanoforecast-v05-competitive-time-series-forecasting-through
- when-directional-accuracy-lies-a-base-rate-honest-benchmark-
Nemotron-H
Mentioned in 2 entries
Qwen 3.5 122B
Mentioned in 2 entries
Dream
Mentioned in 2 entries
- machine-unlearning-for-masked-diffusion-language-models
- pace-dllm-elastic-block-decoding-via-confidence-cliff-estima
Qwen 2.5
Mentioned in 2 entries
- i-gave-five-different-local-llms-a-town-they-invented-facebo
- optimizing-an-ollama-qwen25-ai-agent-fixing-search-aggregati
Qwen3.8 Max
Mentioned in 2 entries
- live-and-ready-to-build-thanks-for-having-us-openrouter-open
- some-food-for-thought-when-designing-benchmarks-so-heres-a-l
Qwen 2512
Mentioned in 2 entries
- comparison-of-low-steps-klein-9b-x-z-image-turbo-x-ernie-tur
- rules-of-thumb-for-regularization-dop-in-ai-toolkit-qwen-251
GEN-1
Mentioned in 2 entries
- robotics-startup-generalist-which-released-its-gen-1-model-t
- source-robotics-startup-generalist-which-released-its-gen-1-
Looped Transformer
Mentioned in 2 entries
- nanbeige42-3b-on-apple-silicon-fixing-deployment-bugs-and-de
- simply-stabilizing-the-loop-via-fully-looped-transformer
NVIDIA Nemotron-3 Nano Omni
Mentioned in 2 entries
- read-the-blog-httpswwwtogetheraiblogtogether-ai-brings-nvidi
- try-now-httpswwwtogetheraimodelsnvidia-nemotron-3-nano-omni
Jamba2-Mini
Mentioned in 2 entries
- uncensored-multi-model-releases-jamba2-mini-qwen35-9b-nikusu
- uncensored-multi-model-releases-longcat-flash-lite-with-mtps
Qwen3.5-9B-Nikusui-v1
Mentioned in 2 entries
- uncensored-multi-model-releases-jamba2-mini-qwen35-9b-nikusu
- uncensored-multi-model-releases-longcat-flash-lite-with-mtps
Qwen3.5-27B-Nikusui-v1
Mentioned in 2 entries
- uncensored-multi-model-releases-jamba2-mini-qwen35-9b-nikusu
- uncensored-multi-model-releases-longcat-flash-lite-with-mtps
QWEN
Mentioned in 2 entries
Claude-Opus-4.6
Mentioned in 2 entries
- a-high-performance-125b-model-now-running-locally-on-just-75
- cudahercules-benchmarking-hardware-aware-expert-level-cuda-o
Ornith-1.5-9B
Mentioned in 2 entries
- does-anyone-have-real-experience-with-ornith-15-9b-for-codin
- we-have-q38-35b-at-home-3x-new-ornith-15-released
DeepSeek4
Mentioned in 2 entries
NVIDIA Nemotron-3-Puzzle-75B-A9B
Mentioned in 2 entries
Gemma 2
Mentioned in 2 entries
- compositional-literary-primitives-in-instruction-tuned-llms-
- recurrence-is-not-enough-causally-validating-multilingual-sa
Ling-3.0-flash-Fin
Mentioned in 2 entries
- antling-open-sourced-ling-30-flash-fin-a-finance-enhanced-mo
- this-finance-model-benchmark-card-is-more-useful-for-what-it
OvisOCR2
Mentioned in 2 entries
Qwen3-Next
Mentioned in 2 entries
Florence 2
Mentioned in 2 entries
- comfyui-face-swap-tutorial-fast-clean-results-vfx-artist-dou
- comfyui-face-swap-tutorial-fast-clean-results-vfx-artist-hey
WAN Video
Mentioned in 2 entries
- comfyui-face-swap-tutorial-fast-clean-results-vfx-artist-dou
- comfyui-face-swap-tutorial-fast-clean-results-vfx-artist-hey
Mage-VL
Mentioned in 2 entries
- mage-vl-an-efficient-codec-native-streaming-multimodal-found
- microsoftmage-vl-hugging-face---an-efficient-codec-native-st
FireRedAudio
Mentioned in 2 entries
- fireredaudio-a-general-purpose-audio-language-model-with-dec
- fireredaudio-fireredtts3-by-fireredteam---huggingface
Laguna-S-2.1
Mentioned in 2 entries
Claude 4.5
Mentioned in 2 entries
- all-of-the-ai-models-have-preferred-names-if-you-asked-claud
- entry-level-guide-to-the-use-of-large-language-models-for-me
Qwen3.5-2B
Mentioned in 2 entries
Qwen Reranker
Mentioned in 2 entries
Claude Code Opus 4.7
Mentioned in 2 entries
- automating-ai-research-is-the-next-major-step-in-ai-we-let-c
- claude-code-opus-47-vs-qwen3627b-on-my-own-little-go-agent
Google Gemini
Mentioned in 2 entries
- pentagons-digital-and-ai-chief-cameron-stanley-confirms-the-
- scidraw-6k-a-multilingual-scientific-illustration-dataset-ge
Gemini Nano Banana 2 Lite
Mentioned in 2 entries
- bringing-speed-and-strong-cost-performance-to-the-market-wit
- googles-gemini-omni-flash-and-nano-banana-2-lite-support-sli
Llama 3.3
Mentioned in 2 entries
3D U-Net
Mentioned in 2 entries
- dalight-3d-a-lightweight-3d-u-net-for-brain-tumor-segmentati
- trajectory-agnostic-asteroid-detection-in-tess-with-deep-lea
Qwen 3.6B
Mentioned in 2 entries
- a-ultra-lightweight-mini-agent---zero-framework-and-localoll
- local-models-do-seem-likely-to-create-an-explosion-of-new-us
DeepSeek-V2
Mentioned in 2 entries
- gqla-group-query-latent-attention-for-hardware-adaptive-larg
- through-the-bottleneck-how-multi-head-latent-attention-separ
MedGemma
Mentioned in 2 entries
- evaluation-of-medical-vision-language-models-hulumed-and-med
- from-manuals-to-maintenance-fine-tuning-medgemma-for-multi-m
GGUF
Mentioned in 2 entries
Spark-X2.5-1.7B
Mentioned in 2 entries
Claude Sonnet 3.7
Mentioned in 2 entries
Jalapeño
Mentioned in 2 entries
- lwiai-podcast-255---gemini-37-jalapeo-qwen-38-drones
- we-plan-to-begin-deploying-jalapeo-in-openais-compute-infras
Qwen 3 8B
Mentioned in 2 entries
- benchmarked-qwen-3-8b-llama-31-8b-qwen-25-7b-vs-gpt-4o-on-m4
- whats-currently-the-smartest-llm-to-use-on-8gb-vram-and-16-r
DeepSeek V4-Flash
Mentioned in 2 entries
- artificial-analysis-deepseeks-v4-flash-costs-0141m-input-and
- deepseek-v4-flash-284b-moe-at-33-toks-single-68-toks-aggrega
PatchTST
Mentioned in 2 entries
- on-the-limits-of-univariate-deep-learning-for-significant-wa
- patched-deltanet-token-level-event-driven-memory-for-linear-
Qwen 3.8 Next
Mentioned in 2 entries
- finally-got-qwen-38-next-running-on-my-v100-6gpu-setup-tp2-p
- i-made-a-custom-llamacpp-build-optimized-for-7900xtx-one-or-
Gemma 7B
Mentioned in 2 entries
- all-models-can-use-web-search-im-using-gemma7b
- cross-generational-transfer-of-adversarial-attacks-reveals-n
Robostral Navigate
Mentioned in 2 entries
- announcing-robostral-navigate-our-first-model-for-embodied-n
- mistral-launches-robostral-navigate-a-hardware-agnostic-robo
TontaubeV1
Mentioned in 2 entries
- tontaubev1---open-tts-model-release-for-local-long-form-gene
- tontaubev1-streaming-text-to-speech-with-hierarchical-codec-
Claude Desktop
Mentioned in 2 entries
- introducing-claude-apps-gateway-for-aws
- local-ai-is-about-to-be-competitive-i-will-do-everything-in-
LongCat-Flash-Lite-Sparse
Mentioned in 2 entries
- meituan-just-dropped-longcat-flash-lite-sparse
- uncensored-multi-model-releases-longcat-flash-lite-sparse-wi
Kimi K3 (Max)
Mentioned in 2 entries
- big-update-among-open-weight-models-kimi-k3-max-is-1-in-the-
- top-15-models-in-agent-arena-vs-median-task-cost-whos-delive
5.6 Sol
Mentioned in 2 entries
- my-feed-has-been-hijacked-this-week-by-frontier-model-influe
- weve-benchmarked-all-the-latest-frontier-models-on-hard-docu
GPT OSS 120B
Mentioned in 2 entries
Bard
Mentioned in 2 entries
- do-yall-remember-the-snake-model-evaluation-test
- from-this-perspective-gemini-is-also-worse-than-the-original
Solar Open 2
Mentioned in 2 entries
qwen4exp
Mentioned in 2 entries
deepseek-v4-flash-0731
Mentioned in 2 entries
- deepseek-v4-flash-0731---surprisingly-usable
- idea-for-a-deepseek-v4-flash-0731-backed-automated-research-
Gemini 2.5 Flash
Mentioned in 2 entries
- i-benchmarked-gemma4e4b-vs-gemma327b-vs-gpt-4o-mini-vs-gemin
- we-used-hflow-to-evaluate-the-latest-open-weights-vlms-for-p
AdamW
Mentioned in 2 entries
- benchmarking-optimizers-for-mlps-in-tabular-deep-learning
- ringmaster-lmo-asynchronous-linear-minimization-oracle-momen
Ornith-1.5
Mentioned in 2 entries
Xiaomi-Robotics-1
Mentioned in 2 entries
- xiaomi-robotics-1-new-robotics-model-released
- xiaomi-robotics-1-scaling-vision-language-action-models-with
DSpark
Mentioned in 2 entries
glm-5.3-flash
Mentioned in 2 entries
- deepseek-v41-flash-vs-glm-53-flash-old-subscription
- i-subscribed-to-ollama-pro-today-deepseek-v4-flash0731-works
Ministral 3 3B
Mentioned in 2 entries
- ministral-3-3b-on-a-galaxy-s21-relayed-a-conversation-betwee
- the-model-punches-above-its-weight-outperforming-gemma-4-e2b
MedSAM3
Mentioned in 2 entries
- a-few-cases-are-all-you-need-an-empirical-study-of-annotatio
- medsam3-delving-into-segment-anything-with-medical-concepts
GPT-Image
Mentioned in 2 entries
Qwen-Omni
Mentioned in 2 entries
DS4
Mentioned in 2 entries
- probably-the-best-way-to-run-ds4-flash-on-a-mac-right-now-19
- qwen38-flash-next-256k-context-16toks-on-ddr4-and-a-tesla-t4
Qwen3-32B
Mentioned in 2 entries
- intertemporal-preference-steering-in-qwen3-via-contrastive-a
- operating-multi-node-full-fine-tuning-on-nvidia-b300-a-field
SmolVLA
Mentioned in 2 entries
- mamba-based-selective-state-space-modeling-improves-the-accu
- when-faster-vla-deployment-changes-closed-loop-behavior-task
PI AGENT
Mentioned in 2 entries
OPENCODE
Mentioned in 2 entries
Qwen3.6-35b
Mentioned in 2 entries
NVIDIA Nemotron 3 Nano Omni
Mentioned in 2 entries
- nvidia-nemotron-3-nano-omni-model-now-available-on-amazon-sa
- nvidia-nemotron-3-nano-omni-powers-multimodal-agent-reasonin
GPT-5.4 nano
Mentioned in 2 entries
- new-paper-worth-reading-gpt-54-nano-plus-a-critic-comparator
- real-time-llm-guardrails-with-jev-comparing-latency-and-cost
Llama Guard
Mentioned in 2 entries
- reflect-guard-enhancing-llm-safeguards-against-adversarial-p
- style-over-substance-content-invariant-wrappers-flip-llm-saf
Qwen3.8-Next-Flash
Mentioned in 2 entries
- installing-6-gpus-in-a-standard-case-rather-than-using-an-op
- qwen38-next-flash-up-to-240ts-on-single-rtx-6000-pro
Qwen 3.6-27b
Mentioned in 2 entries
Qwen 3.5 27B
Mentioned in 2 entries
Qwen 3.8:27B
Mentioned in 2 entries
- benchmarking-qwen3827b-by-having-it-port-a-34k-line-c-game-t
- new-qwen3827b-on-a-39k-line-c-to-single-file-html-threejs-po
ConTextTab
Mentioned in 2 entries
- benchmarking-attention-for-tabular-foundation-models
- tfm-retouche-a-lightweight-input-space-adapter-for-tabular-f
Mitra
Mentioned in 2 entries
- benchmarking-attention-for-tabular-foundation-models
- tfm-retouche-a-lightweight-input-space-adapter-for-tabular-f
Nemotron 3 Nano
Mentioned in 2 entries
- going-from-3b7b-dense-to-nemotron-3-nano-hybrid-mamba-moe-fo
- minimax-m3-no-longer-free-on-ollama-cloud
Mixture of Experts
Mentioned in 2 entries
- application-of-a-mixture-of-experts-based-foundation-model-t
- robustness-of-mixtures-of-experts-to-feature-noise
Claude Opus 5 Max
Mentioned in 2 entries
LLM-JEPA
Mentioned in 2 entries
- dllm-jepa-joint-embedding-predictive-architectures-for-maske
- er-jepa-experience-replay-improves-joint-embedding-predictiv
Qwen 3.5 27b
Mentioned in 2 entries
State-Space Models
Mentioned in 2 entries
- a-benchmark-of-state-space-models-vs-transformers-and-bilstm
- beyond-semantics-how-temporal-biases-shape-retrieval-in-tran
Qwen 3.5-9B
Mentioned in 2 entries
- cloudflare-debuts-open-weight-multimodal-decision-models-cle
- we-tested-out-anthropics-j-lens-probe-at-two-open-models-kim
27B
Mentioned in 2 entries
- 5400-ebay-8x-v100-server-cranks-on-flash-next
- ninfer-and-a-5090-with-38-27b-is-making-me-cry-tears-of-joy-
Stable Diffusion 2.0
Mentioned in 2 entries
Gemma 4:26b
Mentioned in 2 entries
Orukeet
Mentioned in 2 entries
Deepseek V4.1
Mentioned in 2 entries
- cpu-only-experimental-sloppy-deepseek-v41-flash
- how-is-anyone-getting-anything-done-with-ollama-cloud
Qwen-Next
Mentioned in 2 entries
Conditional Variational Autoencoder
Mentioned in 2 entries
- adjoint-based-shape-optimization-of-a-ship-hull-using-a-cond
- synthetic-electric-vehicle-charging-session-generation-using
Q4_K_XL
Mentioned in 2 entries
- need-help-fixing-gemma-4-26b-a4b-toolcalls-in-pi
- r9v-update-now-100-toks-in-tg-on-qwen38-flash-next-iq4xs-on-
Intern-S2-397B
Mentioned in 2 entries
Audio Spectrogram Transformer
Mentioned in 2 entries
- from-masking-to-merging-rethinking-specaugment-for-efficient
- xsq-ast-an-explainable-audio-spectrogram-transformer-framewo
Kimi K2
Mentioned in 2 entries
Bonsai 2
Mentioned in 2 entries
- prism-ml-bonsai-2-joins-our-qwen38-quantization-comparison
- ukisai-swift-series-27b-flash-next-and-bonsai-2-gsq-rco--634
MiniCPM5-2B
Mentioned in 2 entries
- can-current-minicpm5-2bany-best-sota-under-10b-modern-harnes
- minicpm5-2b-vs-spark-x25-4b--64-thinking-x15-speed-while-kee
Koa
Mentioned in 2 entries
- at-dreamforce-salesforce-announces-koa-a-specialized-crm-rea
- salesforce-debuts-koa-a-specialized-model-built-to-reason-ab
DeltaNet
Mentioned in 2 entries
- constant-memory-recall-learned-associations-in-a-fixed-matri
- spectralshift-effective-context-window-extension-of-gated-de
Qwen3.8-Flash-Next-GSQ-RCO-GGUF
Mentioned in 2 entries
Flash Next
Mentioned in 2 entries
- for-web-game-development-i-swear-flash-next-on-q2-is-better-
- ukisai-swift-series-27b-flash-next-and-bonsai-2-gsq-rco--634
Swift-Qwen3.8-27B
Mentioned in 2 entries
- cut-qwen38-27b-reasoning-tokens-by-40----38-thinkingcap-benc
- ukisai-swift-qwen38-27b--583-thinking-x195-speed-while-keepi
deepseek-v4.1-flash
Mentioned in 2 entries
- deepseek-v41-flash-vs-glm-53-flash-old-subscription
- over-the-last-few-hours-some-requests-to-the-deepseek-v41-fl
Qwen 3.8 flash
Mentioned in 2 entries
Claude Enterprise
Mentioned in 2 entries
Ming-Image-0.1-Design
Mentioned in 2 entries
- new-6b-image-model-coming-antling-just-open-sourced-the-ming
- open-source-ming-image-01-design-family-ming-image-01-design
Ming-Image-0.1-Design-Layer
Mentioned in 2 entries
- new-6b-image-model-coming-antling-just-open-sourced-the-ming
- open-source-ming-image-01-design-family-ming-image-01-design
glance-qwen3-vl-4b
Mentioned in 2 entries
- hot-dog-or-not-at-04-sec-jev-style-decision-model-for-image-
- this-is-exactly-what-i-needed-running-glance-qwen3-vl-4b-loc
NV-Reason-CT
Mentioned in 2 entries
- introducing-nv-reason-ct-open-3d-ct-vlm-for-radiologist-chai
- nv-reason-ct-3d-visual-language-model-for-ct-analysis
Faster R-CNN
Mentioned in 2 entries
- difta-3d-depth-consistent-instance-level-feature-transfer-an
- teep-rcnn-texture-enhanced-edge-aware-perception-for-steel-s
System Two
Mentioned in 2 entries
- own-your-harness-folks-i-feel-like-system-one-models-take-cu
- this-is-brilliant-recommended-reading-its-a-great-example-of
Qwen2.1
Mentioned in 2 entries
YuE2
Mentioned in 2 entries
- yue2-ksampler-16x-faster-fix-for-long-songs-and-other-findin
- yue2-unifying-symbolic-and-audio-music-generation-at-frontie
Nemotron 3 Diarization
Mentioned in 2 entries
AliceAI-Foundation-80B-A3B-Base
Mentioned in 2 entries
- im-trying-to-post-train-aliceai-foundation-80b-a3b-base-myse
- yandexaliceai-foundation-80b-a3b-base-russian-developed-comp
Gestalt
Mentioned in 2 entries
GPT-6.1 Astra
Mentioned in 2 entries
- openai-scraps-plans-to-publicly-launch-a-model-dubbed-gpt-61
- openais-gpt-61-sol-delivers-astra-like-performance-at-a-dram
Ember-1
Mentioned in 2 entries
- a-little-ember-1-tldr-seems-like-a-great-model-was-recently-
- independent-numbers-from-valsai-fewer-tokens-per-turn-means-
Clef-flash
Mentioned in 2 entries
- cloudflare-debuts-open-weight-multimodal-decision-models-cle
- introducing-clef-our-open-source-decision-models-and-new-rl-
Strands Decider 2B
Mentioned in 2 entries
- aws-debuts-strands-decider-2b-a-first-lightweight-decision-m
- strands-labs-awss-experimental-agent-development-project-unv
Grok Bot
Mentioned in 2 entries
- a-little-bit-of-codex-in-my-life-a-little-bit-of-claude-code
- what-the-fck-spacexai-is-using-grok-bot-to-build-grok-bot-it
pplx-decider-v1-27b
Mentioned in 2 entries
- alert-again-perplexity-also-just-dropped-a-jev-alternative-o
- pplx-decider-v1-27b-is-fine-tuned-from-qwen38-27b-with-a-250
AliceAI-80B-A3B
Mentioned in 2 entries
- update-2-post-training-yandexaliceai-80b-a3b-instruct-from-s
- update-3-post-training-yandexaliceai-80b-a3b-instruct-from-s
GLM5-Next
Mentioned in 2 entries
PLaMo-2
Mentioned in 2 entries
PLaMo-3
Mentioned in 2 entries
Qwen3.8-Flash-Next 177B
Mentioned in 2 entries