Knowledge catalogue
arXiv:2607.27712v1 Announce Type: new Abstract: Standard masked-language-model fine-tuning applies a uniform masking probability across every token position, assuming reconstruction difficulty is posi
We created a document OCR router that can estimate the complexity of every single page and parse it with the relevant mode 💫 * Some pages are full of native text, which can be directly handled with Li
We have an idea for dinner 2 already :) But what ideas are you all interested in? - technical topics like rl envs, continual learning, cloud agents, world models - general startup / company building f
I can run the mediums, but sometimes I want a faster option that's smarter than Qwen 27B/35B. On my hardware I get like 500 to 800 tok/s prefill and 16 to 22 tok/s gen on ~120B class models, which is
arXiv:2607.26935v1 Announce Type: new Abstract: Bot detectors deployed at scale treat traffic as binary: human or bot. This assumption breaks when AI agents browse the web through browser automation,
arXiv:2607.28148v1 Announce Type: new Abstract: Deep learning has shown promise for automated tongue diagnosis in traditional Chinese medicine (TCM), yet the design space remains underexplored. We con
At Google, AI is a soup-to-nuts endeavor. Obviously, we make leading AI models like Gemini and Nano Banana. We incorporate AI into the tools you use every day (think Gmail, BigQuery, AlloyDB, Google C
I've spent the last couple of days trying different setups (Ollama, Continue, Claude Code, Gemini CLI, OpenRouter...) and at this point I feel like I've spent more time configuring tools than actually
arXiv:2607.16169v3 Announce Type: replace Abstract: Muon is competitive with AdamW in large-scale pre-training, but its operating regime in reinforcement-learning post-training remains unclear. We map
arXiv:2607.26621v2 Announce Type: cross Abstract: Large language models (LLMs) have demonstrated strong reasoning capabilities, motivating their adoption as backbones for foundation recommendation mod
Okay, hear me out. Why is it that every time a new model comes out, all the tests I see are 'make a car game,' 'make a website,' or something equally generic, usually from a prompt that's barely a lin
arXiv:2607.28399v1 Announce Type: new Abstract: Computer-use agents often fail on transient GUI events because they produce the correct action only after the relevant window has already closed. We ide
https://preview.redd.it/vxl4zslewigh1.png?width=1435&format=png&auto=webp&s=5a419870ca0cb13076be6c9ff4ef33d8177d5eda New version is 25% better than previous one and is near GLM-5.2 quality submitted b
I was surprised to see that Deepseek V4 Flash is extremely smart and small enough to fit in setup that can be built with < $50,000. Expensive, but not a datacenter. So I wanted to see the trend over t
arXiv:2607.28478v1 Announce Type: new Abstract: As large language models (LLMs) continue to advance in complex reasoning tasks, they have learned to heavily prioritize explicit conditions provided in
arXiv:2607.27308v1 Announce Type: new Abstract: We introduce ZUNA1.1, a 380M-parameter diffusion autoencoder for flexible EEG signal reconstruction. ZUNA1.1 is capable of reconstructing variable lengt
I've been benchmarking a two-card box for a few weeks and I still can't quite get over some of these numbers, so I'm dumping them here. Box: RTX 4090 (24GB) + RTX 5060 Ti (16GB), i9-13900K, 64GB DDR5.
arXiv:2601.11007v2 Announce Type: replace-cross Abstract: LLM role-playing aims to portray arbitrary characters in interactive narratives, yet existing systems often suffer from limited immersion and
Advancing the price-performance frontier with GPT‑5.6 Huge price drop from OpenAI today: GPT-5.6 Terra got a 20% reduction, and GPT-5.6 Luna got a massive 80% drop. OpenAI credit 5.6 Sol with enabling
“AI Meltdown” is the scenario where the agent goes off the rails and forgets/ignores all previous instructions and soft guardrails. No prompt injection or malicious actor is needed for this to happen.
arXiv:2607.26317v1 Announce Type: cross Abstract: Psychometric calibration for educational tests typically requires costly human response data. Large language models (LLMs) simulated examinees offer a
arXiv:2607.26860v1 Announce Type: new Abstract: We propose amortized moment matching, utilizing neural networks to learn data moments as distributional training signals. By casting diffusion denoisers
And Grok 4.6 comes out in a week BREAKING: Grok 4.5 ranked #1 on LaurenBench with a score of 56.9%, ahead of Claude Sonnet 5, GLM 5.2, Claude Opus 5, Kimi K3 and GPT-5.6. The benchmark tests real-worl
Anthropic: Anthropic says it discovered three of its models had breached three organizations after launching a review in response to the OpenAI-Hugging Face incident — In a review of our cybersecurity
arXiv:2607.27189v1 Announce Type: new Abstract: We introduce APEX-Accounting, a benchmark built by Mercor in partnership with Ramp, to assess whether frontier models can do the real work of accountant
arXiv:2510.20535v2 Announce Type: replace Abstract: Recent techniques such as retrieval-augmented generation or chain-of-thought reasoning have led to longer contexts and increased inference costs. Co
arXiv:2607.26063v1 Announce Type: cross Abstract: Personalised learning systems often assume that mathematical ability is combined of discrete abilities, acquired sequentially and dependent upon first
@ArtificialAnlys ok @openai gets it https://x.com/OpenAI/status/2082878156483219672 We are committed to pushing the model frontier across cost efficiency, capability, and speed. Starting today, we are
arXiv:2607.26344v1 Announce Type: new Abstract: A gradient-based GNN explainer given a molecule with two chemically equivalent nitro groups assigns them attribution scores that are equal to the last b
mimo2: address MTP review feedback (#26228) Co-authored-by: tnhnyc 115956684+tnhnyc@users.noreply.github.com Website: https://llama.app macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silicon (arm6
ggml : Fix issue with kleidiai ci and stringop overflow warning (#26277) Signed-off-by: Jonathan Clohessy Jonathan.Clohessy@arm.com Website: https://llama.app macOS/iOS: macOS Apple Silicon (arm64) ma
metal: fix memory unwire if model is freed without any GPU operations (#26082) metal: fix memory leak if model is freed without any GPU operations metal: run dummy work only if residency sets are used
Remove custom cpu op from the M3 graph, express with stock ops (#26297) Website: https://llama.app macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED macOS I
sync : ggml Website: https://llama.app macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED macOS Intel (x64) iOS XCFramework Linux: Ubuntu x64 (CPU) Ubuntu ar
ggml-cuda: Allow transpose-free gemmv computation (#26171) When matrix's weights are shaped 1xK is leverage a transpose-free computation to use mat_mul_vec_f. Website: https://llama.app macOS/iOS: mac
tests : avoid building get-model.cpp many times (#26317) tests : remove get-model.cpp tests : fix quant type selection Website: https://llama.app macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Sil
llama-context : sync pending async copies before clearing embd_seq (#25676) Website: https://llama.app macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED mac
Test support for alternative conv layout (#25617) add bool cwhn = true to conv_2d test cases add layout check at graph building time extend layout checks for conv2d.cu kernel in CPU back-end kernel ne
vulkan: Support quantized concat (#25684) Website: https://llama.app macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED macOS Intel (x64) iOS XCFramework Lin
server: support inp embd to generate next token (#26313) server: support embd for sampled token fix ~server_batch() Website: https://llama.app macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silico
Written by: Kelli Vanderlee, Stuart Carrera For years, the cybersecurity industry's understanding of software supply chain compromise has been anchored by a few watershed events, including Russian cyb
arXiv:2607.27023v1 Announce Type: new Abstract: Evaluating large generative models across benchmarks is time-consuming and computationally expensive. This drives the need for methods that can estimate
I recently shared the original MindControl PoC (and on github) - sampler-level guided reasoning budgets for llama.cpp, nudging the model with self-aware statements about its own thinking budget instea
arXiv:2607.26247v1 Announce Type: new Abstract: Low-rank adaptation (LoRA) fine-tunes large pretrained models at a fraction of the cost of full fine-tuning, but its performance depends strongly on how
arXiv:2607.26232v1 Announce Type: new Abstract: Background manipulation is a practical but under-specified image-forensics setting: the manipulated evidence can sit outside the salient foreground obje
arXiv:2606.19651v2 Announce Type: replace-cross Abstract: Three-dimensional (3D) brain MRI is central to clinical neurology and neuro-oncology, where generative models could augment under-represented
arXiv:2607.26828v1 Announce Type: new Abstract: Large language models increasingly support scientific and algorithmic discovery through inference-time search over evaluated candidates. Existing adapti
arXiv:2603.24800v2 Announce Type: replace Abstract: In this paper, we uncover the hidden potential of Diffusion Transformers (DiTs) to significantly enhance generative tasks. Through an in-depth analy
arXiv:2607.26200v1 Announce Type: new Abstract: Content-moderation classifiers are usually evaluated in isolation, but deployment requires choosing where to intervene and what follows a flag. We evalu
arXiv:2607.26470v1 Announce Type: new Abstract: Multi-turn information-seeking conversations require both multi-hop reasoning and long-range dependency tracking across turns. However, existing RAG sys
arXiv:2607.27054v1 Announce Type: new Abstract: Knowledge distillation (KD) enables a compact student model to learn from a powerful teacher and has become an effective paradigm for model compression.
ComfyUI Face Swap Tutorial: Fast, Clean Results VFX artist Doug Hogan walks through a fully automated face swap workflow built inside ComfyUI - combining Florence 2, SAM2, and WAN Video into a single
Consistent with what I found with Qwen3.6 a while back: Claude Code uses 2-3x as many tokens than (many) other harnesses at similar success rate. - Unoptimized? - Buggy? - Deliberate (coz that helps i
arXiv:2607.27143v1 Announce Type: new Abstract: High-stakes decision systems in credit scoring, fraud detection, healthcare, and industrial safety require reliable uncertainty quantification under sev
arXiv:2607.26952v1 Announce Type: new Abstract: We introduce CreditCardQA, the first financial literacy benchmark for numerical reasoning derived from real credit card agreements. The dataset contains
arXiv:2607.26792v1 Announce Type: cross Abstract: Probabilistic K-line forecasting describes uncertainty in four complementary prices, namely open--high--low--close (OHLC). However, it introduces two
arXiv:2607.26596v1 Announce Type: new Abstract: Multimodal large language models (MLLMs) have demonstrated remarkable capabilities by integrating visual and textual understanding within a unified tran
arXiv:2607.26933v1 Announce Type: cross Abstract: Federated Learning (FL) is vulnerable to backdoor attacks because of its distributed nature in edge computing scenarios. Existing defense methods show
arXiv:2607.27178v1 Announce Type: new Abstract: State-of-the-art retrieval models increasingly rely on closed training data, creating a reproducibility gap. We present an open end-to-end recipe for tr
arXiv:2607.26368v1 Announce Type: new Abstract: Financial disclosures contain numerical claims, temporal statements, entity references, policy commitments, and risk descriptions that may conflict in q