Knowledge catalogue
arXiv:2608.03467v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR) com- monly optimizes each correct completion as an independent learning signal. In GRPO, this comp
arXiv:2608.03649v1 Announce Type: new Abstract: Fewer visual tokens do not guarantee lower end-to-end latency. We evaluate break-even with a reproducible protocol that accounts for decision overhead,
arXiv:2608.03506v1 Announce Type: new Abstract: Self-consistency assumes the most frequent answer among sampled reasoning traces is the most reliable, but this can fail in causal reasoning: samples of
arXiv:2608.01679v2 Announce Type: replace Abstract: Persistent memory allows (self-evolving) LLM agents to adapt across tasks by consolidating heterogeneous interaction histories into reusable facts,
arXiv:2608.03722v1 Announce Type: new Abstract: Collective intelligence research treats disagreement as evidence of epistemic diversity: if agents express different views, the group should retain capa
arXiv:2608.03201v1 Announce Type: new Abstract: Safety guards are widely used to filter harmful content and are typically trained via supervised fine-tuning on labeled prompt-response pairs. We audit
arXiv:2606.15673v2 Announce Type: replace Abstract: Web agents act through long interaction sequences, yet existing benchmarks evaluate only terminal success, discarding all process information and of
arXiv:2608.02829v1 Announce Type: new Abstract: Model families train every size from scratch. Can a pretrained large model be converted into a smaller sibling? We characterize the 1.4B->410M conversio
arXiv:2608.04008v1 Announce Type: new Abstract: Benchmarks that measure the forecasting ability of large language models are almost always retrospective: the event has happened, the answer is somewher
Xiaomi-Robotics-1 is a robot foundation model trained on over 100K hours of real-world manipulation trajectories. It is a Vision-Language-Action (VLA) model engineered for out-of-the-box mobile manipu
Gary Marcus noted that investor Dwarkesh predicts Anthropic will generate more than $100 billion in revenue this year, according to a source. A chart accompanying the claim omitted DeepSeek’s pricing
arXiv:2608.02657v1 Announce Type: cross Abstract: Agentic LLMs are vulnerable to indirect prompt injection (IPI) attacks, e.g., malicious side-tasks hidden in external tool results. While many efforts
Liquid AI released LFM2.5-2.6B today, and this might be more relevant to local AI than another massive model most people cannot run. The model is only 2.69B parameters, has 128K context, supports tool
arXiv:2608.01258v1 Announce Type: new Abstract: The realism of images generated by multimodal large language models (MLLMs), such as GPT Image2 and Nano Banana2, has improved rapidly in recent years.
arXiv:2608.02062v1 Announce Type: cross Abstract: Faster-than-Nyquist signaling improves spectral ef- ficiency by deliberately introducing inter-symbol interference. Classical sequence detectors such
arXiv:2608.00218v1 Announce Type: new Abstract: Agentic LLMs exhibit three consequential tool-use failures: invalid arguments (validity), unnecessary calls (over-calling), and omitted calls when tools
arXiv:2608.02300v1 Announce Type: new Abstract: Existing astronomy foundation models provide strong galaxy representations, but adapting them to new survey conditions and survey-specific morphology re
arXiv:2606.15974v2 Announce Type: replace Abstract: Despite the significant advancement of LLMs in conversation summarization, their evaluation remains limited by insufficient scenarios, input lengths
A new llama.cpp PR (#26563) adds a heatmap that tracks which MoE experts are used most often. Instead of keeping every expert on the GPU or offloading all of them, it caches the frequently selected ex
arXiv:2608.02539v1 Announce Type: cross Abstract: We present a simple Gaussian approximation to the finite-sample distribution of the classical ridge regression estimator. Our approximation captures t
arXiv:2608.00705v1 Announce Type: cross Abstract: Reported verdicts on GraphRAG versus vector RAG disagree, and the evidence is typically tied to a single corpus, embedder, and judge -- and, we show,
arXiv:2608.02505v1 Announce Type: cross Abstract: Can scientific abduction occur without continuous sensorimotor embodiment? Recent arguments in AI and philosophy of science hold that genuine hypothes
arXiv:2512.08163v2 Announce Type: replace Abstract: Deep neural networks (DNNs) are increasingly used as functional models of human vision, yet standard monocular depth estimation (MDE) benchmarks lar
arXiv:2608.00434v1 Announce Type: new Abstract: Multi-Token Prediction (MTP) has emerged as an effective paradigm that augments a shared Large Language Model backbone with auxiliary heads, training th
arXiv:2608.00850v1 Announce Type: new Abstract: Physics-informed neural networks (PINNs) have emerged as a versatile approach for solving nonlinear partial differential equations (PDEs), yet achieving
arXiv:2608.02049v1 Announce Type: cross Abstract: Bosonic quantum systems provide a hardware-efficient platform for quantum information processing but remain challenging to characterise due to their l
arXiv:2509.24560v2 Announce Type: replace Abstract: Extended Chain-of-Thought (CoT) reasoning has significantly bolstered the capabilities of medical large language models (LLMs). However, current mod
arXiv:2608.00832v1 Announce Type: new Abstract: Structured plan-generation agents are often evaluated as if a plan has quality in isolation, yet many realistic planning tasks require asking how a cand
arXiv:2608.00702v1 Announce Type: new Abstract: Nighttime aerial image enhancement is challenged by spatially nonuniform exposure, mixed illumination, and weak structural evidence, while registered no
arXiv:2608.00101v1 Announce Type: cross Abstract: AI coding agents like GitHub Copilot, Claude Code, and Codex interleave multi-step LLM inference with tool execution, creating a workload different fr
arXiv:2608.00009v1 Announce Type: new Abstract: Long-term memory remains a critical bottleneck for conversational AI agents, whose finite context windows cannot support coherent recall across thousand
All models in the Portal are discounted by 20 %, except GPT‑5.6 Terra and Luna (50 % off) and DeepSeek V4 Flash 0731 (90 % off). The latest Alibaba Qwen release, Qwen3.8‑Max, is now available for Herm
arXiv:2512.04954v3 Announce Type: replace Abstract: We present a novel technique for amortized posterior estimation using Normalizing Flows trained with likelihood-weighted importance sampling. This a
arXiv:2608.02248v1 Announce Type: cross Abstract: Deformable image registration (DIR) is a core problem in medical image analysis; but, unlike labeling decision problems such as classification and seg
arXiv:2608.00737v1 Announce Type: new Abstract: Hard-constrained recurrent physics-informed networks (HRPINNs) embed known dynamics inside a recurrent numerical integrator and restrict a neural branch
arXiv:2608.01865v1 Announce Type: new Abstract: Automatic speech recognition (ASR) performance degrades sharply on dysarthric speech, yet how disordered articulation reshapes a model's internal repres
arXiv:2608.00881v1 Announce Type: new Abstract: Large language model agents increasingly act through stateful tools, yet model generation and environment execution remain serialized at every step. As
Appreciate it! High performance, low cost. Try it out today. 💻 Qwen 3.8 Max is available on AI Gateway. It scores mid-pack on the DeepsecBench cybersecurity leaderboard at one of the lowest costs per
arXiv:2608.01291v1 Announce Type: new Abstract: We present ArabicDialectSafety, a human-curated Arabic safety dataset of 25,071 prompts covering six Arabic varieties: Modern Standard Arabic, Syrian, E
arXiv:2608.01336v1 Announce Type: new Abstract: Modern autonomous-driving fleets record far more video than human reviewers can inspect. This motivates the need for an automatic clip triage mechanism,
arXiv:2608.00652v1 Announce Type: new Abstract: Human assistance in robotics spans around several tasks such as navigation, object manipulation, and placement, where a key challenge is selecting targe
arXiv:2608.00693v1 Announce Type: new Abstract: Schema linking is a critical component of Text-to-SQL systems, but existing approaches often trade off contextual modeling capacity, score-based control
arXiv:2608.00515v1 Announce Type: cross Abstract: Large language model tutors can be correct and helpful yet disclose an answer or decisive reasoning before that disclosure is authorized. We formalize
arXiv:2608.00701v1 Announce Type: cross Abstract: Reweighting source samples to match a target covariate distribution is a standard response to distribution shift when generalizing evidence from one p
arXiv:2608.02602v1 Announce Type: new Abstract: Language remains an outlier in generative modeling: while images, video, and audio are increasingly modeled in continuous latent spaces, text generation
arXiv:2608.01935v1 Announce Type: new Abstract: Prior work in Ancient Greek NLP relies on corpora that do not disambiguate the phonemic vowel length of alpha, iota, and ypsilon, together known as the
arXiv:2608.00145v1 Announce Type: cross Abstract: Accurate estimation of left ventricular (LV) orientation is essential for cardiac magnetic resonance (CMR) imaging and downstream analysis. Existing m
Amazon Web Services Inc. today launched Kiro Crew, an autonomous workspace that keeps artificial intelligence coding agents running all day and all night. The company said it developed Crew to allow d
ggml: use dynamic allocation for split graph inputs (#22789) ggml: use dynamic allocation for split graph inputs Replace fixed-size GGML_SCHED_MAX_SPLIT_INPUTS arrays with dynamically allocated buffer
vocab : validate default special token ids (#26506) Website: https://llama.app macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED macOS Intel (x64) iOS XCFra
server: add get_info tool (#26522) server: add get_info tool fix --rpc in docs server: harden get_info probe result handling Report the OS as unknown when the probe process fails to spawn or times out
tests: add model resolution test on synthetic repo listings (#26172) tests: add model resolution test on synthetic repo listings Include download.cpp and arg.cpp inside a namespace with hf_cache monke
model : support MTP in GLM-4.7-Flash (#24868) Website: https://llama.app macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED macOS Intel (x64) iOS XCFramework
vendor : update BoringSSL to 0.20260803.0 (#26523) Website: https://llama.app macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED macOS Intel (x64) iOS XCFram
vendor : update cpp-httplib to 0.52.0 (#26485) Website: https://llama.app macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED macOS Intel (x64) iOS XCFramewor
chat : add new template for DeepSeek V4 Flash 0731 (#26398) common/chat: update DeepSeek V4 templates Align the DeepSeek V4 templates with the official encoders while keeping parser behavior out of th
Extended SYCL oneDNN SDPA to non-FP16 KV caches (Q4_0–Q8_0 and FP32) (#25874) sycl: extend oneDNN SDPA to Q4_0-Q8_0 and F32 KV caches Extends the oneDNN SDPA path (PR #25222) to handle non-F16 KV cach
sycl: parallelize the non-contiguous concat kernel (#25852) sycl: parallelize the non-contiguous concat kernel Launch geometry only: the non-contiguous concat kernel launched a single-lane work-group
llama : move n_vocab from llama_sampler_data to penalty_sampler (#26520) This matches how it is done for logit_bias and mirostat samplers, see #25262 (comment) Website: https://llama.app macOS/iOS: ma
model : allow reshape of tensors during load (#26531) Website: https://llama.app macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED macOS Intel (x64) iOS XCF