Search: “reasoning”

SEABench: Benchmarking Endogenous Misalignment In Self-Evolving Agents

arXiv:2609.35596v2 Announce Type: replace-cross Abstract: Self-evolving LLM agents have gained prominence for their ability to improve after deployment by modifying their harness, including their cont

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Semantic Misalignment in Vision-Language Models under Perceptual Degradation

DGX agent

arXiv:2601.08355v3 Announce Type: replace Abstract: Vision-Language Models (VLMs) are increasingly deployed in autonomous driving and embodied AI systems, where reliable perception is critical for saf

safetyarxiv-cs-cv

1 Oct 2026

Sequential Bayesian Evaluation of Large Language Model Behavior

model-releasesarxiv-cs-cl

arXiv:2511.10661v2 Announce Type: replace Abstract: It is increasingly important to evaluate the characteristics of systems based on large language models (LLMs). Evaluations in this context often rel

SETA: Scaling Environments for Terminal Agents

arXiv:2607.10891v2 Announce Type: replace Abstract: Large language models (LLMs) are rapidly shifting toward agents that solve tasks through diverse interfaces, including web and graphical user interf

Smaller Models, Better Rejects: Preference Distillation Scaling

arXiv:2609.38987v1 Announce Type: new Abstract: Preference distillation typically treats a teacher response as preferred and the student's own response as rejected. This assumes that self-generated fa

safetyarxiv-cs-lg

Speculative Safety Honeypot: Toward Proactive Defense Against Multi-turn Agent Attacks

arXiv:2609.39549v1 Announce Type: cross Abstract: As Large Language Model (LLM) agents are increasingly deployed in complex environments, multi-turn interaction attacks have become a significant secur

Tutorials

SPOON: Towards Coherent Compositional 3D Scene Generation from Uncalibrated Multi-view Images

arXiv:2609.39590v1 Announce Type: new Abstract: Compositional 3D scene generation aims to recover complete 3D object shapes and their spatial arrangement from visual observations. Recent image-conditi

tutorialsarxiv-cs-cv

Tutorials

TACIT: Optimization Models that Learn from Their Mistakes

arXiv:2609.38434v1 Announce Type: cross Abstract: Real-world optimization problems are difficult to model accurately because many objectives and constraints reside in domain experts' tacit knowledge,

tutorialsarxiv-cs-ai

Talk2Agent: Benchmarking Voice Interfaces for Text Agents

arXiv:2609.38867v1 Announce Type: new Abstract: Large language model (LLM) computer-use agents are typically evaluated with clean written instructions, despite speech being an increasingly popular int

The critical distinction between base LLMs (2024 and earlier) and modern LRMs is not symbolic tool use. It's the switch from a transductive …

model-releasesfrancois-chollet--x

The critical distinction between base LLMs (2024 and earlier) and modern LRMs is not symbolic tool use. It's the switch from a transductive paradigm (intuit the answer to the query) to an inductive pa

Today we’re introducing Extract v2.5 - a series of frontier agents tuned for document extraction. The agents (cost-effective, agentic, agent…

Today we’re introducing Extract v2.5 - a series of frontier agents tuned for document extraction. The agents (cost-effective, agentic, agentic plus) are tuned for value accuracy and grounding. Our ext

agentsjerry-liu--x

Typographic Attack Against VLM-based AI-generated Image Detection

arXiv:2609.39662v1 Announce Type: new Abstract: Vision-language models (VLMs) are increasingly used for AI-generated image (AIGI) detection, providing natural-language explanations for authenticity ju

researcharxiv-cs-cv

Uncovering Uncontrolled Repetition through Residual Stream Dynamics

arXiv:2609.38802v1 Announce Type: cross Abstract: Uncontrolled repetition can prolong autoregressive generation in large language models (LLMs) and enable resource consumption attacks. Prior analyses

researcharxiv-cs-ai

Understanding as No-Arbitrage: Bounded Dutch Books as a Definition and Training Objective for Language Models

arXiv:2609.39341v1 Announce Type: new Abstract: Does a language model merely predict tokens, or does it understand what it says? We make this question measurable by defining 'understanding' through th

Update #2: Post training yandex/AliceAI-80B-A3B [instruct!] from scratch

model-releasesr-localllama

Last update: https://www.reddit.com/r/LocalLLaMA/comments/1wu9ksu/update_yandexaliceai_80ba3b_fine_tune_progress/ - basically, an instruct fine tune on the base model using a synthetic distilled data

When Context Changes: Understanding Update Failures in LLMs

arXiv:2609.38866v1 Announce Type: new Abstract: As preferences, goals, and facts change, LLM agents must use the current state while earlier versions remain in context. Yet they can answer with an old

When Scientific Contradictions Are Lost in Translation

arXiv:2609.38621v1 Announce Type: new Abstract: Two scientific findings can disagree without contradicting each other. Determining whether they conflict requires knowing whether they describe comparab

Which Models Work Well Together? Measuring Heterogeneity for LLM Team Selection

arXiv:2609.38274v1 Announce Type: new Abstract: The performance ceiling of an LLM team is constrained not only by individual model capabilities, but also by inter-member error resonance and predictive

researcharxiv-cs-cl

Who Owns That? Evaluating Ownership Intuitions in Large Language Models

arXiv:2609.39483v1 Announce Type: new Abstract: Ownership establishes rights over the use, control, and transfer of objects. Understanding these relations is essential for AI systems to interact appro

researcharxiv-cs-ai

Working Around the Compute Ceiling: Byte-Exact Memory in Galahad Makes LLM Reading a One-Time Cost LLM Reading a One-Time Cost

arXiv:2609.39358v1 Announce Type: cross Abstract: A transformer language model performs a bounded amount of computation per token, and recent work by Vishal Sikka, former CEO of Infosys, argues that t

World-as-Graph: Relational World Modeling Through Latent Space Graphs

arXiv:2609.38927v1 Announce Type: cross Abstract: World models aim to learn representations of real-world environments and predict their future evolution. Recent object-centric world models have made

safetyarxiv-cs-cv

You're Hired: Strategic Model Selection for LLM Collaboration

arXiv:2609.38816v1 Announce Type: new Abstract: While multi-agent and model collaboration algorithms gain traction to combine the strengths of diverse Large Language Models (LLMs), existing systems re

agentsarxiv-cs-cl

Absorbed in Inertia: Activation Analysis for Computer-Use Agents

arXiv:2609.37176v1 Announce Type: new Abstract: Computer-use agents have become increasingly capable of executing tasks on live desktops through natural-language instructions, based on trajectories of

agentsarxiv-cs-ai

Agentic Federated Learning: Rule-Based Client and Server Agents for Adaptive Training

arXiv:2609.35914v1 Announce Type: cross Abstract: Federated Learning (FL) enables collaborative model training across distributed clients without sharing raw data, making it suitable for privacy-sensi

agentsarxiv-cs-lg

An interview with OpenAI Chief Research Officer Mark Chen on the Hugging Face incident, slowing AI development, shifting 5%-10% of compute to safety, and more (Will Douglas Heaven/MIT Technology Review)

Will Douglas Heaven / MIT Technology Review: An interview with OpenAI Chief Research Officer Mark Chen on the Hugging Face incident, slowing AI development, shifting 5%-10% of compute to safety, and m

safetytechmeme

BAAI/AREX-2 - 27B - Agent model based on Qwen3.8 27B

model-releasesr-localllama

'AREX-2 is a 27B-parameter long-horizon agent model from the Beijing Academy of Artificial Intelligence (BAAI). It learns to improve a solution over multiple test-time rounds: propose, measure, reflec

Bathtubs, Boundaries, and Sandboxes: AI Regulatory Learning under Legal Uncertainty

arXiv:2601.04094v4 Announce Type: cross Abstract: Effective regulation of AI is a defining policy challenge, driven by their integration into all aspects of society. To remain responsive to their rapi

Better Nearest Neighbor Graph Indices via (Efficient) LLM-Guided Pruning

arXiv:2609.36359v1 Announce Type: new Abstract: Graph-based approximate nearest neighbor search (ANNS) is widely used for large-scale semantic search. Its indices are constructed primarily based on ge

researcharxiv-cs-ai

Beyond Pixels: A Vector-to-Graph Framework for Reliable Schematic Auditing

arXiv:2602.11678v2 Announce Type: replace Abstract: Multimodal Large Language Models (MLLMs) have shown remarkable progress in visual understanding, yet they suffer from a critical limitation: structu

Beyond Rule-Based Mutation Testing: Test-Aware Mutant Generation Using Large Language Models

arXiv:2609.35841v1 Announce Type: cross Abstract: Mutation testing evaluates test-suite adequacy by injecting synthetic faults into program code. However, traditional rule-based tools often generate l

Causal-EVC: Breaking Emotional Spurious Causality via Spatiotemporal Grounding and Counterfactual Intervention

model-releasesarxiv-cs-cv

arXiv:2609.36776v1 Announce Type: new Abstract: Emotional Video Captioning aims to generate factually accurate and emotionally empathetic descriptions. While recent methods have recognized the importa

Cognitive Expert Language Models Better Align with the Corresponding Brain Systems

arXiv:2609.36239v1 Announce Type: new Abstract: Large language models (LLMs) can predict human brain activity across a variety of brain regions during natural language comprehension. Typically, howeve

safetyarxiv-cs-cl

Local Ai

ComfyUI-TypedDecision: run Jev-style decision models locally in ComfyUI

local-air-stablediffusion

Tasks where an MLLM doesn't write text but only makes a decision, which started with Jev, seem to be getting popular. (This is sometimes called 'typed decision'.) I thought it would be fun to use insi

COMiT: Learning Structured Visual Tokens through Sequential Communication

arXiv:2602.20731v2 Announce Type: replace-cross Abstract: Discrete image tokenizers provide a sequential interface for vision and multimodal models, but are typically optimized for reconstruction or c

Tutorials

Corpus-Guided Dual-Path Propagation for Graph Retrieval-Augmented Generation

arXiv:2609.37661v1 Announce Type: new Abstract: Graph-based retrieval-augmented generation supports multi-hop retrieval by organizing corpus information into graphs. However, existing relation-free gr

tutorialsarxiv-cs-cl

Counterfactual Probing for Parallel Unmasking with Hidden Forest Structure

model-releasesarxiv-cs-lg

arXiv:2609.37841v1 Announce Type: new Abstract: Masked generative models offer parallel token prediction, but accurate parallel sampling must account for dependencies among tokens. When dependencies a

CounterSteer: Suppressing Indirect Prompt Injection with Activation Steering

arXiv:2609.36570v1 Announce Type: cross Abstract: Indirect prompt injection makes an LLM agent treat untrusted retrieved text as instructions. We present CounterSteer, an inference-time defense that s

CypherTurn: A Multi-Turn Benchmark for Conversational Text-to-Cypher Evaluation and the Autonomy Divergence

model-releasesarxiv-cs-cl

arXiv:2609.36987v1 Announce Type: new Abstract: Graph databases are increasingly queried through natural language, yet every existing benchmark evaluates isolated single-turn queries rather than the m

DeepRewind: Predicting and Repairing Premature Commitments in Deep Research Agents

arXiv:2609.36344v1 Announce Type: new Abstract: Deep-research agents conduct long-horizon investigations through iterative search, evidence evaluation, belief revision, and synthesis. However, they ma

agentsarxiv-cs-cl

Distilling What Matters: Confidence-Aware Selective Distillation for Large Language Models

arXiv:2609.36734v1 Announce Type: new Abstract: Knowledge Distillation (KD) trains a smaller-capacity student model to imitate a larger-capacity teacher model by matching output distributions, implici

safetyarxiv-cs-cl

Dr. OPD: Learning What to Follow for Optimal On-Policy Distillation of Large Language Models

arXiv:2609.38025v1 Announce Type: cross Abstract: On-policy distillation (OPD) trains a student on its own generated responses using dense, token-level supervision from a stronger teacher. Vanilla OPD

DSWM: Decomposed Spatio-Temporal World Model for Demand-Driven UAV Base Station Repositioning

arXiv:2609.36845v1 Announce Type: cross Abstract: Uncrewed aerial vehicle base stations (UAV-BSs) are expected to cover traffic demand that shifts across space and time, yet most repositioning schemes

agentsarxiv-cs-ai

Dual-Channel Robust Group-Relative Policy Optimization via Advantage and Sequence-Weight Estimation

arXiv:2609.36944v1 Announce Type: new Abstract: Group-relative policy optimization relies on reward-derived advantages and sequence-level likelihood weights, both of which can be sensitive to localize

Dyad: Extending Large Language Models with Native Typed Decision-Making

arXiv:2609.36116v1 Announce Type: new Abstract: We study how to build more capable general-purpose agents by extending large language models (LLMs) with native typed decision-making. We introduce Dyad

safetyarxiv-cs-lg

EasyPPO: Stabilizing the Critic Is Key

arXiv:2609.36802v1 Announce Type: new Abstract: A key strength of Proximal Policy Optimization (PPO) is its learned critic, which uses historical trajectories collected during reinforcement learning t

safetyarxiv-cs-lg

Engineering Simplicity: Simple Mechanism Interfaces Steer LLM Agents

arXiv:2609.36365v1 Announce Type: new Abstract: Can interaction formats and textual scaffolds help large language model (LLM) agents make better decisions, and do better decisions come with better exp

EngiWorld: What Can Frontier Agents Deliver in Professional Engineering Environments?

arXiv:2609.37686v1 Announce Type: new Abstract: Autonomous agents have made rapid progress in general-purpose computer use, but reliable automation of professional industrial engineering remains out o

Evaluating and Benchmarking the System One Model Jev

arXiv:2609.37647v1 Announce Type: cross Abstract: Jev is a commercial System One model from TypeSafe AI that does not generate text: given a state and typed questions, it returns a choice from fixed o