Search: “reasoning”

Boosting RL-Based Visual Reasoning with Selective Adversarial Entropy Intervention

arXiv:2512.10414v2 Announce Type: replace Abstract: Recently, reinforcement learning (RL) has become a common choice in enhancing the reasoning capabilities of vision-language models (VLMs). Consideri

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Model Releases

CultureForest: Understanding and Evaluating Cultural Norm Grounded Reasoning in LLMs

DGX agent

arXiv:2606.01879v1 Announce Type: new Abstract: Existing research largely reduces cultural intelligence in LLMs to a knowledge-level problem, overlooking whether models can effectively utilize their a

model-releasesarxiv-cs-cl

2 Jun 2026

Decomposed On-Policy Distillation for Vision-Language Reasoning: Steering Gradients for Visual Grounding

arXiv:2606.00564v1 Announce Type: cross Abstract: While on-policy distillation offers dense supervision for training small reasoning models, its optimization dynamics in the multimodal domain remain u

safetyarxiv-cs-cl

On Information Self-Locking in Reinforcement Learning for Active Reasoning of LLM agents

arXiv:2603.12109v2 Announce Type: replace Abstract: Reinforcement learning (RL) has become a de facto paradigm for building LLM-based agents that act, interact, and reason over extended task horizons.

agentsarxiv-cs-ai

PLanAR: Planning-Language-Grounded Agentic Reasoning for Robot Manipulation

arXiv:2602.01662v4 Announce Type: replace Abstract: Recent advances in vision-language models (VLMs) have enabled increasing progress in real-world robot manipulation. However, long-horizon manipulati

agentsarxiv-cs-ro

Plausibility Is Not Prediction: Contrastive Evidence for LLM-Based Cellular Perturbation Reasoning

arXiv:2606.01042v1 Announce Type: cross Abstract: Perturbation experiments are central to understanding cellular mechanisms, but remain costly and sparse, motivating prediction of gene expression resp

R^3: Composed Video Retrieval via Reasoning-Guided Recalling and Re-ranking

arXiv:2606.01113v1 Announce Type: new Abstract: The CoVR-R challenge evaluates composed video retrieval, where a system must retrieve a target video from a large gallery given a reference video and a

researcharxiv-cs-cv

Hardware

Skill-Based Mixture-of-Experts: Adaptive Routing for Heterogeneous Reasoning via Inferred Skills

arXiv:2503.05641v4 Announce Type: replace-cross Abstract: Combining existing pre-trained LLMs is a promising approach for diverse reasoning tasks. However, task-level expert selection is often too coa

hardwarearxiv-cs-ai

TechGraphRAG: An Agentic Graph-Augmented RAG Framework for Technical Literature Reasoning

arXiv:2606.01613v1 Announce Type: cross Abstract: This paper presents an agentic retrieval-augmented generation (RAG) framework for domain-specific technical reasoning support, instantiated over a cur

agentsarxiv-cs-ai

TLG: Temporal-Logic Grounding for Video Question Answering via Source-Annotation Reconstruction and Category-Targeted Reasoning

model-releasesarxiv-cs-cv

arXiv:2606.01591v1 Announce Type: new Abstract: The TimeLogic Challenge evaluates formal temporal-logic reasoning over video - 16 operators (before, after, until, since, always, co-occur, ordering, ..

Training-Free Composed Video Retrieval via Visual Representation-Guided Video-LLM Reasoning

arXiv:2606.02321v1 Announce Type: new Abstract: Recent advances in large vision-language models have expanded video retrieval from simple text-based search to more flexible scenarios, where users may

researcharxiv-cs-cv

TRON: Targeted Rule-Verifiable Online Environments for Visual Reasoning RL

arXiv:2606.01599v1 Announce Type: new Abstract: Reinforcement learning (RL) for visual reasoning needs scalable, verifiable, and controllable training signals. Existing visual RL post-training trains

Hardware

Develop Physical AI Reasoning, World, and Action Models with NVIDIA Cosmos 3

NVIDIA Cosmos 3 is a frontier foundation model for physical AI that combines physical reasoning, world generation, and action generation within a single open model. The model uses a Mixture-of-Transfo

hardwarenvidia-developer

Applications

Do Large Language Models Encode Institutional Experience? Evidence from Cross-Linguistic Moral Reasoning Under Ambiguity

arXiv:2605.30934v1 Announce Type: cross Abstract: Large language models (LLMs) exhibit systematic differences in moral reasoning across languages, yet the source of this variation remains unclear. We

applicationsarxiv-cs-ai

FAM-Bench: A Multimodal Benchmark for Condition-Aware Food-as-Medicine Reasoning

arXiv:2605.31410v1 Announce Type: new Abstract: Food-as-Medicine requires models to reason beyond what a dish is or what nutrition it contains: they must decide whether a concrete food choice is appro

Rationalize: Shared Semantic Reasoning for Human-AI Alignment

arXiv:2605.30632v1 Announce Type: cross Abstract: We introduce Rationalize, a role-pair framework for shared semantic reasoning between humans and AI models in data-driven sensemaking. Building on ide

Tutorials

Reliable Self-Improvement Training by Verifying Reasoning, Not Just Answers

arXiv:2603.21558v2 Announce Type: replace Abstract: Self-improvement training, where models learn from self-generated solutions, promises sustained capability gains but suffers from a pervasive failur

tutorialsarxiv-cs-ai

Industry

Sources: at Build, Microsoft plans to unveil a Copilot 'super app', a new reasoning AI model developed by Microsoft AI, and lots of Windows developer features (Tom Warren/The Verge)

Tom Warren / The Verge: Sources: at Build, Microsoft plans to unveil a Copilot “super app”, a new reasoning AI model developed by Microsoft AI, and lots of Windows developer features — Build will incl

industrytechmeme

Beyond 3D VQAs: Injecting 3D Spatial Priors into Vision-Language Models for Enhanced Geometric Reasoning

arXiv:2605.30231v1 Announce Type: cross Abstract: Vision-Language Models (VLMs) often struggle with robust 3D spatial reasoning. Prevailing methods that rely on fine-tuning with 3D visual question-ans

From Meta-Thought to Execution: Cognitively Aligned Post-Training for Generalizable and Reliable LLM Reasoning

arXiv:2601.21909v2 Announce Type: replace Abstract: Current LLM post-training methods optimize complete reasoning trajectories through Supervised Fine-Tuning (SFT) followed by outcome-based Reinforcem

Tutorials

Knowing What to Solve Before How: Preplan Empowered LLM Mathematical Reasoning

arXiv:2605.30245v1 Announce Type: new Abstract: Current plan-based reasoning methods improve large language models (LLMs) by inserting a planning stage before execution, giving rise to the question ri

tutorialsarxiv-cs-cl

Local Ai

MediHive: A Decentralized Agent Collective for Medical Reasoning

arXiv:2603.27150v2 Announce Type: replace Abstract: Large language models (LLMs) have revolutionized medical reasoning tasks, yet single-agent systems often falter on complex, interdisciplinary proble

local-aiarxiv-cs-ai

Mining or Synthesis? Rethinking Exploration Efficiency in Iterative Alignment of Mathematical Reasoning

arXiv:2602.05370v3 Announce Type: replace Abstract: Iterative Direct Preference Optimization (DPO) has emerged as a widely used paradigm for aligning Large Language Models on reasoning tasks. Existing

safetyarxiv-cs-cl

OmniMatBench: A Human-Calibrated Multimodal Reasoning Benchmark Across 19 Materials Science Subfields

arXiv:2605.29833v1 Announce Type: new Abstract: As multimodal language models play an increasingly important role in scientific research, materials science offers a critical testbed due to its interdi

BenGER: Benchmarking LLM Systems on Subsumption-Based Legal Reasoning in German Law

arXiv:2605.28183v1 Announce Type: cross Abstract: We introduce the BenGER (Benchmark for German Law) dataset for evaluating LLM systems on subsumption-based legal reasoning in German law. The BenGER d

FABSVer: Faster Training and Better Self-Verification for LLM Mathematical Reasoning

arXiv:2605.28389v1 Announce Type: new Abstract: While large language models have made significant progress in mathematical reasoning, they remain unreliable at judging the correctness of their own sol

safetyarxiv-cs-cl

Applications

FinBoardBench: Benchmarking Dynamic Wealth Management and Strategic Financial Reasoning of LLMs via Board Game Simulations

arXiv:2605.27896v1 Announce Type: new Abstract: Recently, large language models (LLMs) have achieved superior performance in static financial reasoning and simple dynamic trading tasks. However, exist

applicationsarxiv-cs-cl

ForestHG-Trace: Traceable Long-Horizon Ecological Reasoning over Large-Scale Forest Scenes

model-releasesarxiv-cs-cv

arXiv:2605.27590v1 Announce Type: new Abstract: Remote sensing question answering (RS-QA) often requires more than direct semantic prediction, especially in large-scale forest scenes where ecological

MACReD: A Multi-Agent Collaborative Reasoning Framework for Reaction Diagram Parsing

arXiv:2605.28077v1 Announce Type: new Abstract: Parsing chemical reaction diagrams from scientific literature is challenging due to heterogeneous layouts, intertwined visual elements, and the difficul

Skill-Conditioned Gated Self-Distillation for LLM Reasoning

arXiv:2605.28791v1 Announce Type: cross Abstract: On-policy self-distillation (SD) improves LLM reasoning by using teacher-side privileged information (PI) to turn sparse verifier outcomes into dense

TRACER: Turn-level Regret Matching with Inner Reinforcement Credit for Cooperative Multi-LLM Reasoning

arXiv:2605.28699v1 Announce Type: new Abstract: Large language models increasingly rely on either reinforcement learning or multi-agent prompting to improve reasoning, yet these two paradigms remain d

ARBITER: Reasoning Trajectory Basins and Majority Vote Failures in Test-Time Sampling

model-releasesarxiv-cs-lg

arXiv:2605.26172v1 Announce Type: new Abstract: When language models use test-time sampling, they generate multiple reasoning trajectories and select an answer by majority vote. We show that these tra

Athena: Enhancing Multimodal Reasoning with Data-efficient Process Reward Models

arXiv:2506.09532v5 Announce Type: replace-cross Abstract: We present Athena-PRM, a multimodal process reward model (PRM) designed to evaluate the reward score for each step in solving complex reasonin

Composition Collapse: Stable Factual Knowledge Does Not Imply Compositional Reasoning

arXiv:2605.26789v1 Announce Type: new Abstract: Post-training is routinely evaluated through aggregate benchmark scores that treat multi-hop reasoning as a single capability -- as if a model that answ

Learning to Adapt SFT Data for Better Reasoning Generalization

arXiv:2605.26924v1 Announce Type: new Abstract: Large language models (LLMs) have achieved remarkable progress, with post-training playing a crucial role in enhancing their reasoning capabilities. Amo

SOLE-R1: Video-Language Reasoning as the Sole Reward for On-Robot Reinforcement Learning

model-releasesarxiv-cs-cl

arXiv:2603.28730v2 Announce Type: replace-cross Abstract: Vision-language models (VLMs) have shown impressive capabilities across diverse tasks, motivating efforts to leverage these models to supervis

Uncertainty-Aware Budget Allocation for Adaptive Test-Time Reasoning

arXiv:2605.26849v1 Announce Type: new Abstract: Sampling multiple responses improves language model reasoning, but uniform compute allocation is inefficient: easy questions are over-sampled while hard

Local Ai

Understanding Reasoning in LLMs through Strategic Information Allocation under Uncertainty

arXiv:2603.15500v2 Announce Type: replace Abstract: LLMs often exhibit Aha moments such as self-correction after tokens like 'Wait,' yet the underlying mechanism remains unclear. Standard LLMs collaps

local-aiarxiv-cs-ai

AgentFugue: Agent Scaling for Long-Horizon Tasks through Collective Reasoning

arXiv:2605.24486v1 Announce Type: new Abstract: Recent progress on long-horizon agentic tasks has been driven largely by scaling up individual agents through stronger models, better tools, and more ef

agentsarxiv-cs-ai

DUEL: Adversarial Self-Play for Multimodal Reasoning

arXiv:2605.24794v1 Announce Type: cross Abstract: Reinforcement learning (RL) has emerged as an effective paradigm for improving the reasoning capability of vision-language models (VLMs). However, RL-

FloorplanQA: A Benchmark for Spatial Reasoning in LLMs using Structured Representations

arXiv:2507.07644v4 Announce Type: replace Abstract: We introduce FloorplanQA, a diagnostic benchmark for evaluating spatial reasoning in large language models (LLMs). FloorplanQA is grounded in struct

HiMed: Incentivizing Hindi Reasoning in Medical LLMs

model-releasesarxiv-cs-cl

arXiv:2605.24635v1 Announce Type: new Abstract: Medical large language models hold promise for reducing healthcare disparities, yet Hindi remains severely underrepresented. While medical LLMs excel in

Pragmatic Reasoning improves LLM Code Generation

arXiv:2502.15835v5 Announce Type: replace-cross Abstract: Pragmatic reasoning helps interlocutors infer intended meaning from ambiguous or underspecified messages by considering shared context and cou

Residual Drift Dominates Contradiction in Multi-Turn Constraint Reasoning

arXiv:2605.23940v1 Announce Type: new Abstract: How do multi-turn reasoning systems fail? The expected answer is logical contradiction, in which the system's maintained state becomes unsatisfiable. We

Convex Compositional Reasoning Models

arXiv:2605.23395v1 Announce Type: new Abstract: Compositional energy-based models can generalize to larger combinatorial reasoning problems by reusing a learned factor energy across many local constra

researcharxiv-cs-lg

Decoding the Critique Mechanism in Large Reasoning Models

arXiv:2603.16331v2 Announce Type: replace Abstract: Large Reasoning Models (LRMs) exhibit backtracking and self-verification mechanisms that enable them to revise intermediate steps and reach correct

researcharxiv-cs-lg

Evaluating Counterfactual Strategic Reasoning in Large Language Models

arXiv:2603.19167v2 Announce Type: replace Abstract: We evaluate Large Language Models (LLMs) in repeated game-theoretic settings to assess whether strategic performance reflects genuine reasoning or r

VI-CuRL: Stabilizing Verifier-Independent RL Reasoning via Confidence-Guided Variance Reduction

arXiv:2602.12579v2 Announce Type: replace-cross Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as a dominant paradigm for enhancing Large Language Models (LLMs) reasoning,