Papers

New on arXiv

Thu 9 Oct 2025

516 papers

Towards Neurocognitive-Inspired Intelligence: From AI's Structural Mimicry to Human-Like Functional CognitionDeep reinforcement learning methods → Deep RL algorithms and representations
MEDEQUALQA: Evaluating Biases in LLMs with Counterfactual ReasoningLLM reasoning → Clinical and biomedical applications
Energy-Guided Diffusion Sampling for Long-Term User Behavior Prediction in Reinforcement Learning-based RecommendationDiffusion models → Diffusion recommenders
A Multi-Component Reward Function with Policy Gradient for Automated Feature Selection with Dynamic Regularization and Bias MitigationDeep reinforcement learning methods → Deep RL algorithms and representations
VisualDAN: Exposing Vulnerabilities in VLMs with Visual-Driven DAN CommandsMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness +1
Kelp: A Streaming Safeguard for Large Models via Latent Dynamics-Guided Risk DetectionLLM safety and security → Safety alignment, refusal and guardrails
When Search Goes Wrong: Red-Teaming Web-Augmented Large Language ModelsLLM safety and security → Prompt injection and agent security
Pattern Enhanced Multi-Turn Jailbreaking: Exploiting Structural Vulnerabilities in Large Language ModelsLLM safety and security → Jailbreaking
Time-Aware Feature Selection: Adaptive Temporal Masking for Stable Sparse Autoencoder TrainingMechanistic interpretability → SAE features and their applications
CDE: Concept-Driven Exploration for Reinforcement LearningDeep reinforcement learning methods → RL for robot locomotion and manipulation
FOLK: Fast Open-Vocabulary 3D Instance Segmentation via Label-guided Knowledge DistillationKnowledge distillation → Detection and dense prediction distillation
What Is Your Agent's GPA? A Framework for Evaluating Agent Goal-Plan-Action AlignmentLLM agents and tool use → General LLM agent frameworks and benchmarks
Reinforcement Learning-Driven Edge Management for Reliable Multi-view 3D ReconstructionDeep reinforcement learning methods → RL for energy, networks and operations
CommandSans: Securing AI Agents with Surgical Precision Prompt SanitizationLLM agents and tool use → Agent threats, guardrails and governance +1
Search-on-Graph: Iterative Informed Navigation for Large Language Model Reasoning on Knowledge GraphsLLM reasoning → Knowledge- and tool-augmented reasoning
D-CoDe: Scaling Image-Pretrained VLMs to Video via Dynamic Compression and Question DecompositionMultimodal LLMs / vision-language models → Visual token pruning and efficiency
MOSAIC: Multi-agent Orchestration for Task-Intelligent Scientific CodingLLM agents and tool use → Web, search, and research agents
Benchmarking Chinese Commonsense Reasoning with a Multi-hop Reasoning PerspectiveLLM reasoning → Knowledge- and tool-augmented reasoning
COMPASS: Enhancing Agent Long-Horizon Reasoning with Evolving ContextLLM reasoning → Reasoning for agents, planning and code +1
Q-Router: Agentic Video Quality Assessment with Expert Model Routing and Artifact LocalizationLLM agents and tool use → Multimodal and embodied agents
Geometry-aware Policy ImitationOffline RL and imitation learning from data → Robot manipulation imitation learning
Zero-Shot Policy Transfer in Reinforcement Learning using Buckingham's Pi TheoremDeep reinforcement learning methods → Deep RL algorithms and representations
Dissecting Embodied Abilities in Multimodal Language Models through Skill-level Evaluation and DiagnosisMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
Graph Diffusion Transformers are In-Context Molecular DesignersDiffusion models → Flow matching, time series and science
When to Reason: Semantic Router for vLLMLLM reasoning → Test-time compute and efficient thinking
Towards Unified World Models for Visual Navigation via Memory-Augmented Planning and ForesightWorld models → World models for navigation
Thinking Longer, Not Always Smarter: Evaluating LLM Capabilities in Hierarchical Legal ReasoningLLM reasoning → Reasoning for agents, planning and code
BigCodeArena: Unveiling More Reliable Human Preferences in Code Generation via ExecutionLLM alignment / RLHF / preference optimisation → Process reward models for reasoning
Don't Waste Mistakes: Leveraging Negative RL-Groups via Confidence ReweightingLLM reasoning → Reinforcement learning for reasoning
FreqCa: Accelerating Diffusion Models via Frequency-Aware CachingDiffusion models → Diffusion foundations and sampling
RAG4Tickets: AI-Powered Ticket Resolution via Retrieval-Augmented Generation on JIRA and GitHub DataRetrieval-augmented generation → RAG for code
dInfer: An Efficient Inference Framework for Diffusion Language ModelsDiffusion models → Discrete and text diffusion
RA-Gen: A Controllable Code Generation Framework Using ReAct for Multi-Agent Task ExecutionLLM agents and tool use → Software engineering coding agents
Inverse-Free Wilson Loops for Transformers: A Practical Diagnostic for Invariance and Order SensitivityLLM reasoning → Prompting and chain-of-thought reasoning
Can Reasoning Path still be Effective as Input? Bridging Post-Reasoning to Chain-of-Thought CompressionLLM reasoning → Test-time compute and efficient thinking
Mitigating Over-Refusal in Aligned Large Language Models via Inference-Time Activation EnergyLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks +1
Automating Android Build Repair: Bridging the Reasoning-Execution Gap in LLM Agents with Domain-Specific ToolsLLM agents and tool use → Software engineering coding agents
BLAZER: Bootstrapping LLM-based Manipulation Agents with Zero-Shot Data GenerationLLM agents and tool use → Multimodal and embodied agents +1
Scalable Offline Metrics for Autonomous DrivingEnd-to-end autonomous driving → End-to-end system architecture
MATRIX: Multimodal Agent Tuning for Robust Tool-Use ReasoningLLM reasoning → Multimodal reasoning +2
NaViL: Rethinking Scaling Properties of Native Multimodal Large Language Models under Data ConstraintsLLM reasoning → Multimodal reasoning +1
How to Teach Large Multimodal Models New SkillsMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
ResAD: Normalized Residual Trajectory Modeling for End-to-End Autonomous DrivingEnd-to-end autonomous driving → End-to-end system architecture
SciVideoBench: Benchmarking Scientific Video Reasoning in Large Multimodal ModelsMultimodal LLMs / vision-language models → Video understanding with MLLMs
Improving Reasoning for Diffusion Language Models via Group Diffusion Policy OptimizationLLM reasoning → Diffusion reasoning and test-time scaling +1
Dream to Recall: Imagination-Guided Experience Retrieval for Memory-Persistent Vision-and-Language NavigationWorld models → World models for navigation
MM-HELIX: Boosting Multimodal Long-Chain Reflective Reasoning with Holistic Platform and Adaptive Hybrid Policy OptimizationLLM reasoning → Multimodal reasoning +1
Permutation-Invariant Spectral Learning via Dyson DiffusionDiffusion models → Flow matching, time series and science
SpatialLadder: Progressive Training for Spatial Reasoning in Vision-Language ModelsLLM reasoning → Spatial and embodied reasoning +1
X2Video: Adapting Diffusion Models for Multimodal Controllable Neural Video RenderingDiffusion models → Video generation
CoMAS: Co-Evolving Multi-Agent Systems via Interaction RewardsLLM agents and tool use → General LLM agent frameworks and benchmarks
Convergence Theorems for Entropy-Regularized and Distributional Reinforcement LearningDeep reinforcement learning methods → RL theory and sample complexity
Which Heads Matter for Reasoning? RL-Guided KV Cache CompressionLLM reasoning → Test-time compute and efficient thinking +1
DYNAMIX: RL-based Adaptive Batch Size Optimization in Distributed Machine Learning SystemsDeep reinforcement learning methods → RL for energy, networks and operations
CaRT: Teaching LLM Agents to Know When They Know EnoughLLM agents and tool use → General LLM agent frameworks and benchmarks
AutoMLGen: Navigating Fine-Grained Optimization for Coding AgentsLLM agents and tool use → General LLM agent frameworks and benchmarks
To Sink or Not to Sink: Visual Information Pathways in Large Vision-Language ModelsMultimodal LLMs / vision-language models → Visual token pruning and efficiency
InstructX: Towards Unified Visual Editing with MLLM GuidanceMultimodal LLMs / vision-language models → Image editing and quality assessment +1
DeepPrune: Parallel Scaling without Inter-trace RedundancyLLM reasoning → Test-time compute and efficient thinking
The Visual Iconicity Challenge: Evaluating Vision-Language Models on Sign Language Form-Meaning MappingMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Video-STAR: Reinforcing Open-Vocabulary Action Recognition with ToolsMultimodal LLMs / vision-language models → Video understanding with MLLMs
Looking to Learn: Token-wise Dynamic Gating for Low-Resource Vision-Language ModellingMultimodal LLMs / vision-language models → Visual token pruning and efficiency
Don't Run with Scissors: Pruning Breaks VLA Models but They Can Be RecoveredVision-language-action models → Efficient VLA inference and adaptation
Wavefunction Flows: Efficient Quantum Simulation of Continuous Flow ModelsDiffusion models → Diffusion foundations and sampling
ARES: Multimodal Adaptive Reasoning via Difficulty-Aware Token-Level Entropy ShapingLLM reasoning → Test-time compute and efficient thinking
Gaze on the Prize: Shaping Visual Attention with Return-Guided Contrastive LearningDeep reinforcement learning methods → RL training of language and agent models
Large Scale Diffusion Distillation via Score-Regularized Continuous-Time ConsistencyDiffusion models → Diffusion foundations and sampling
ClauseLens: Clause-Grounded, CVaR-Constrained Reinforcement Learning for Trustworthy Reinsurance PricingDeep reinforcement learning methods → RL training of language and agent models
Reinforcing Diffusion Models by Direct Group Preference OptimizationLLM reasoning → Reinforcement learning for reasoning +2
Optimal Stopping in Latent Diffusion ModelsDiffusion models → Diffusion foundations and sampling
VideoVerse: Does Your T2V Generator Have World Model Capability to Synthesize Videos?World models → Video-generation world models and simulators
FlyLoRA: Boosting Task Decoupling and Parameter Efficiency via Implicit Rank-Wise Mixture-of-ExpertsMixture-of-experts → Mixture of LoRA experts
QAgent: A modular Search Agent with Interactive Query UnderstandingRetrieval-augmented generation → Agentic RAG, agents and memory +1
Guided Star-Shaped Masked DiffusionDiffusion models → Discrete and text diffusion
Hyperspectral data augmentation with transformer-based diffusion modelsDiffusion models → Inverse problems and image restoration
Evaluating Small Vision-Language Models on Distance-Dependent Traffic PerceptionMultimodal LLMs / vision-language models → Driving VLMs and open-vocabulary detection
AutoRed: A Free-form Adversarial Prompt Generation Framework for Automated Red TeamingLLM safety and security → Prompt injection and agent security
Beyond Pass@k: Breadth-Depth Metrics for Reasoning BoundariesLLM reasoning → Reinforcement learning for reasoning
LinVideo: A Post-Training Framework towards O(n) Attention in Efficient Video GenerationDiffusion models → Video generation
Iterated Agent for Symbolic RegressionLLM agents and tool use → Web, search, and research agents
To Ask or Not to Ask: Learning to Require Human FeedbackLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF
First Try Matters: Revisiting the Role of Reflection in Reasoning ModelsLLM reasoning → Test-time compute and efficient thinking
Bridging the Physics-Data Gap with FNO-Guided Conditional Flow Matching: Designing Inductive Bias through Hierarchical Physical ConstraintsDiffusion models → Flow matching, time series and science
Beyond Turn Limits: Training Deep Search Agents with Dynamic Context WindowLLM reasoning → Knowledge- and tool-augmented reasoning +1
One Stone with Two Birds: A Null-Text-Null Frequency-Aware Diffusion Models for Text-Guided Image InpaintingDiffusion models → 3D, point cloud and scene generation
SViM3D: Stable Video Material Diffusion for Single Image 3D GenerationDiffusion models → 3D, point cloud and scene generation
Evaluation of a Robust Control System in Real-World Cable-Driven Parallel RobotsDeep reinforcement learning methods → RL for robot locomotion and manipulation
Mix- and MoE-DPO: A Variational Inference Approach to Direct Preference OptimizationLLM alignment / RLHF / preference optimisation → LLM preference optimization methods +1
Opponent Shaping in LLM AgentsLLM agents and tool use → Reinforcement learning for LLM agents
Contrastive Decoding for Synthetic Data Generation in Low-Resource Language ModelingSynthetic data and self-improvement for LLMs → LLM synthetic data generation
Simulating Teams with LLM Agents: Interactive 2D Environments for Studying Human-AI DynamicsLLM agents and tool use → Social simulation and multi-agent debate
The Alignment Waltz: Jointly Training Agents to Collaborate for SafetyLLM agents and tool use → Agent threats, guardrails and governance +2
Chain-of-Trigger: An Agentic Backdoor that Paradoxically Enhances Agentic RobustnessLLM agents and tool use → Agent threats, guardrails and governance +1
Enhancing Reasoning for Diffusion LLMs via Distribution Matching Policy OptimizationLLM reasoning → Diffusion reasoning and test-time scaling
UAMDP: Uncertainty-Aware Markov Decision Process for Risk-Constrained Reinforcement Learning from Probabilistic ForecastsDeep reinforcement learning methods → RL theory and sample complexity
Selection, Reflection and Self-Refinement: Revisit Reasoning Tasks via a Causal LensLLM reasoning → Test-time compute and efficient thinking
Expressive Value Learning for Scalable Offline Reinforcement LearningDiffusion models → Diffusion and flow policies for robotics +1
Training-Free Group Relative Policy OptimizationLLM reasoning → Reinforcement learning for reasoning
R-Horizon: How Far Can Your Large Reasoning Model Really Go in Breadth and Depth?LLM reasoning → Test-time compute and efficient thinking
InstructUDrag: Joint Text Instructions and Object Dragging for Interactive Image EditingDiffusion models → Image generation, text-to-image and editing

All 516