model releases

Chip Export rules: made China/Huawei stronger, accomplished relatively little. USG stakes in US AI: will make Mistral and other “sovereign A…

Chip Export rules: made China/Huawei stronger, accomplished relatively little. USG stakes in US AI: will make Mistral and other “sovereign AI” efforts stronger, freak out the rest of the world, escala

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Closing the Loop on Latent Reasoning via Test-Time Reconstruction

arXiv:2606.06252v1 Announce Type: new Abstract: Recent work moves intermediate reasoning from natural-language traces into latent or cache-level representations to reduce token overhead and avoid a di

CogManip: Benchmarking Manipulative Behavior in Multi-Turn Interactions with Large Language Model

arXiv:2606.06099v1 Announce Type: new Abstract: Whether Large Language Models (LLMs) exhibit covert psychological manipulation in complex human-AI interactions has garnered increasing safety concerns.

Critic-Guided Heterogeneous Multi-Agent Reasoning for Reliable Mathematical Problem Solving

arXiv:2606.05704v1 Announce Type: new Abstract: Recent Large Language Models (LLMs) have shown impressive reasoning abilities; but they are still susceptible to hallucinations, intermediate reasoning

Cross-Epoch Adaptive Rollout Optimization for RL Post-Training

arXiv:2606.05606v1 Announce Type: cross Abstract: LLM post-training often relies on reinforcement learning methods that sample multiple rollouts per prompt, yet most existing approaches use a fixed ro

CTIConnect: A Benchmark for Retrieval-Augmented LLMs over Heterogeneous Cyber Threat Intelligence

arXiv:2510.11974v2 Announce Type: replace-cross Abstract: Cyber Threat Intelligence (CTI) is foundational to modern cybersecurity, enabling organizations to proactively defend against evolving threats

Data Flow Control: Data Safety Policies for AI Agents

arXiv:2606.05679v1 Announce Type: cross Abstract: Agents increasingly generate SQL, orchestrate pipelines, and automate data analysis on behalf of users. While recent work improves query correctness,

Do More Agents Help? Controlled and Protocol-Aligned Evaluation of LLM Agent Workflows

arXiv:2606.05670v1 Announce Type: new Abstract: Does adding more agents help an LLM workflow once compared systems share the same benchmark loader, tool access, answer contract, usage accounting, and

DPBench: Structural Determinants of Multi-Agent LLM Coordination Under Simultaneous Resource Contention

arXiv:2602.13255v2 Announce Type: replace Abstract: We present DPBench, a benchmark for evaluating coordination in multi-agent systems built from large language models. Existing benchmarks measure tas

DragOn: A Benchmark and Dataset for Drag-Based GUI Interactions

arXiv:2606.06322v1 Announce Type: new Abstract: GUI agents - vision-based models that control desktops, web browsers, and mobile devices through graphical user interfaces - promise to automate a wide

Edit-R2: Context-Aware Reinforcement Learning for Multi-Turn Image Editing

arXiv:2606.05950v1 Announce Type: new Abstract: Text-guided image editing has advanced rapidly with diffusion models and unified multimodal foundation models. However, most existing methods remain con

Enhancing Software Engineering Through Closed-Loop Memory Optimization

arXiv:2606.05646v1 Announce Type: cross Abstract: Large language models (LLMs) have enabled powerful software engineering (SE) agents capable of navigating complex codebases and resolving real-world i

Evaluating Agentic Configuration Repair for Computer Networks

arXiv:2606.06212v1 Announce Type: new Abstract: Misconfigurations in computer networks remain a major source of critical Internet outages. Research is turning to Large Language Models (LLMs) to automa

Evaluation of LLMs for Mathematical Formalization in Lean

arXiv:2606.05632v1 Announce Type: new Abstract: Within the past few years, the ability of Large Language Models (LLMs) to generate formal mathematical proofs has improved drastically. We provide a com

Exploring LLMs for South Asian Music Understanding and Generation

arXiv:2606.05522v1 Announce Type: cross Abstract: Recent advancements in Large Language Models (LLMs) have shown promising results in music understanding and generation tasks. However, existing works

Fireworks Training Platform keeps expanding. Leading US open weight model Nemotron 3 Ultra is now ready for post-training: SFT and DPO via L…

model-releasesfireworks-ai--x

Fireworks Training Platform keeps expanding. Leading US open weight model Nemotron 3 Ultra is now ready for post-training: SFT and DPO via LoRA or full-parameter, on the same infrastructure that serve

Four papers out recently: 1. http://political-manipulation.ai: Measures and reduces political bias in LLMs; Claude is especially biased 2. h…

model-releasesdan-hendrycks--x

Four papers out recently: 1. http://political-manipulation.ai: Measures and reduces political bias in LLMs; Claude is especially biased 2. http://aibetrayal.com: The public can insert backdoors into A

GenTI: Benchmarking LLMs for Autonomous IDPS Rule Generation for Unseen Attacks

arXiv:2606.05844v1 Announce Type: cross Abstract: Rule-based Intrusion Detection and Prevention Systems (IDPS) offer precise attack detection as well as mitigation, however their manually crafted, sig

Geographic Bias and Diversity in AI Evaluation

arXiv:2606.05187v1 Announce Type: cross Abstract: Among the many challenges hindering the responsible development and deployment of AI, arguably none has faced more intense scrutiny than bias in its v

GITCO: Gated Inference-Time Context Optimization in TSFMs

arXiv:2606.05332v1 Announce Type: new Abstract: Patch-based Time Series Foundation Models (TSFMs) suffer from context poisoning: structurally anomalous patches capture disproportionate attention and s

Goedel-Architect: Streamlining Formal Theorem Proving with Blueprint Generation and Refinement

arXiv:2606.06468v1 Announce Type: new Abstract: We introduce Goedel-Architect, an agentic framework for formal theorem proving in Lean 4 centered on blueprint generation and refinement. A blueprint is

GuardNet: Ensemble Strategies of Shallow Neural Networks for Robust Prompt Injection and Jailbreak Detection

arXiv:2606.05566v1 Announce Type: new Abstract: Large Language Models (LLMs) have transformed natural language processing, but they remain vulnerable to Prompt Injection (PI) and Jailbreak (JB) attack

How Far Did They Go? The Persuasive Tactics of Covert LLM Agents in a Discontinued Field Experiment

arXiv:2606.05256v1 Announce Type: new Abstract: This study analyzes a publicly released dataset from a discontinued field experiment on Reddit's r/ChangeMyView. The intervention, conducted by unknown,

I Know What You Meme, Even If it Emerged Today: Understanding Evolving Memes through Open-World Knowledge Acquisition

arXiv:2606.05316v1 Announce Type: new Abstract: Multimodal memes are dynamic and often require up to date background knowledge for interpretation. Existing methods often overlook such knowledge or rel

I want AI to succeed, and have been positive about many systems (Claude Code, AlphaFold, AlphaGeometry, Cicero, etc). It’s not AI that I hat…

I want AI to succeed, and have been positive about many systems (Claude Code, AlphaFold, AlphaGeometry, Cicero, etc). It’s not AI that I hate; it’s bullshit and greed that I despise. Not my fault that

Introducing Harness-1, a 20B search agent trained with a state-externalizing harness. > frontier-level long-horizon search, rivaling Opus-4.…

model-releasesclem-delangue--x

Introducing Harness-1, a 20B search agent trained with a state-externalizing harness. > frontier-level long-horizon search, rivaling Opus-4.6 and outperforming GPT-5.4 > Context-1-level cost and laten

it’s a bad day when @scaling01 goes all Gary Marcus on the media 🤣

it’s a bad day when @scaling01 goes all Gary Marcus on the media 🤣 This is a good example of how AI news gets distorted. The example is from Tagesschau, the most watched news program in Germany. Anthr

“I've tried to find where this stops being circular and I can't.”

“I've tried to find where this stops being circular and I can't.” 🦔Google signed a deal to pay SpaceX 920 million a month for access to 110,000 Nvidia GPUs at SpaceX data centers. The contract runs Oc

Learning What Matters: Probabilistic Task Selection via Mutual Information for Model Finetuning

arXiv:2507.12612v3 Announce Type: replace-cross Abstract: Supervised fine-tuning performance for large language models depends strongly on how training budget is distributed across a heterogeneous set

LLMCodec: Adapting Video Codecs for Efficient Weight Compression of Large Language Models

arXiv:2606.05861v1 Announce Type: cross Abstract: The rapid development of large language models(LLMs) has led to remarkable advances in natural language processing. However, the increasing scale of t

Memory is Reconstructed, Not Retrieved: Graph Memory for LLM Agents

arXiv:2606.06036v1 Announce Type: new Abstract: Despite recent progress, LLM agents still struggle with reasoning over long interaction histories. While current memory-augmented agents rely on a stati

Minimizing the Hidden Cost of Scales: Graph-Guided Ultra-Low-Bit Quantization for Large Language Models

arXiv:2606.05429v1 Announce Type: new Abstract: Post-training quantization (PTQ) is critical for the efficient deployment of large language models (LLMs). Recent ultra-low-bit PTQ methods rely on rigi

Multilingual Fine-Tuning via Localized Gradient Conflict Resolution

arXiv:2606.05613v1 Announce Type: new Abstract: The rapid evolution of Large Language Models (LLMs) has established cross-lingual versatility as a defining feature of modern systems. However, fine-tun

No Need to Train Your RDB Foundation Model

arXiv:2602.13697v2 Announce Type: replace Abstract: Relational databases (RDBs) contain vast amounts of heterogeneous tabular information that can be exploited for predictive modeling purposes. But si

OPRD: On-Policy Representation Distillation

arXiv:2606.06021v1 Announce Type: cross Abstract: On-policy distillation (OPD) supervises the student only in output space by matching next-token probabilities. This output-only paradigm has two limit

PC Layer: Polynomial Weight Preconditioning for Improving LLM Pre-Training

arXiv:2606.06470v1 Announce Type: cross Abstract: We propose a preconditioning (PC) layer, a weight parameterization via polynomial preconditioner that ensures stable weight conditioning throughout LL

ProfiliTable: Profiling-Driven Tabular Data Processing via Agentic Workflows

arXiv:2605.12376v2 Announce Type: replace Abstract: Table processing-including cleaning, transformation, augmentation, and matching-is a foundational yet error-prone stage in real-world data pipelines

PSEBench: A Controllable and Verifiable Benchmark for Evaluating LLMs in Patient Safety Event Triage

arXiv:2606.05463v1 Announce Type: new Abstract: Patient safety event triage, determining whether a clinical event is reportable under jurisdiction-specific policy, is a high-stakes task typically perf

Retry Policy Gradients in Continuous Action Spaces

arXiv:2606.05888v1 Announce Type: new Abstract: Retry-based objectives such as pass@K and max@K optimize the best return obtained from multiple sampled trajectories, and recent work has shown that the

Reward-Decomposed Reinforcement Learning for Immersive Video Role-Playing

arXiv:2605.04733v2 Announce Type: replace Abstract: Text-based role-playing models can imitate character styles, but often fail to capture scene atmosphere and evolving tension, which are crucial for

Reward Learning through Ranking Mean Squared Error

arXiv:2601.09236v3 Announce Type: replace-cross Abstract: Reward design remains a significant bottleneck in applying reinforcement learning (RL) to real-world problems. A popular alternative is reward

Running Python code in a sandbox with MicroPython and WASM

model-releasessimon-willison

I've been experimenting with different approaches to running code in a sandbox for several years now, but my latest attempt feels like it might finally have all of the characteristics I've been lookin

Safety Paradox: How Enhanced Safety Awareness Leaves LLMs Vulnerable to Posterior Attack

arXiv:2606.05614v1 Announce Type: new Abstract: Large language models (LLMs) are rigorously aligned to refuse harmful requests, a process that inherently cultivates a latent capacity to evaluate and r

SagnacAssisted Enhanced OTDR for Distributed Acoustic Sensing: A Standardized Benchmark and Engineering Evaluation Framework

arXiv:2606.05754v1 Announce Type: cross Abstract: Phase-sensitive optical time-domain reflectometry (phi-OTDR) is widely used in large-scale distributed acoustic sensing (DAS) because it provides dist

SciVisAgentSkills: Design and Evaluation of Agent Skills for Scientific Data Analysis and Visualization

arXiv:2606.05525v1 Announce Type: new Abstract: Recent advances in agentic visualization have enabled the translation of natural language into executable scientific visualization (SciVis) workflows. W

Search-Time Contamination in Deep Research Agents: Measuring Performance Inflation in Public Benchmark Evaluation

arXiv:2606.05241v1 Announce Type: cross Abstract: Public benchmarks enable fair and reproducible evaluation of LLM reasoning, but they become fragile for deep research agents that actively search the

⚠️⚠️ Seismic shift ⚠️⚠️ It’s a good day to be Mistral. Nobody is going to trust an American AI company that is partly owned by the US Govern…

⚠️⚠️ Seismic shift ⚠️⚠️ It’s a good day to be Mistral. Nobody is going to trust an American AI company that is partly owned by the US Government. Just the way the US doesn’t trust Huawei. After this m

Selective-Advantage Entropy-Adaptive Horizon GRPO: Asymmetric Token-Level Discounting for Efficient Reinforcement Learning of Language Models

arXiv:2606.05434v1 Announce Type: cross Abstract: Group Relative Policy Optimisation (GRPO) has emerged as an effective reinforcement-learning algorithm for aligning language models on reasoning tasks