Papers

New on arXiv

Tue 3 Jun 2025

507 papers

SynLang and Symbiotic Epistemology: A Manifesto for Conscious Human-AI CollaborationLLM reasoning → Reasoning benchmarks and evaluation
daDPO: Distribution-Aware DPO for Distilling Conversational AbilitiesLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
PARC: A Quantitative Framework Uncovering the Symmetries within Vision Language ModelsMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
Argus Inspection: Do Multimodal Large Language Models Possess the Eye of Panoptes?LLM reasoning → Multimodal reasoning +1
ss-Mamba: Semantic-Spline Selective State-Space ModelState-space sequence models (S4, Mamba) → Time series forecasting with SSMs
CLAIM: Mitigating Multilingual Object Hallucination in Large Vision-Language Models with Cross-Lingual Attention InterventionMultimodal LLMs / vision-language models → Hallucination mitigation in LVLMs
ChemGraph: An Agentic Framework for Computational Chemistry WorkflowsLLM agents and tool use → Medical and scientific domain agents
From Transformers to Large Language Models: A systematic review of AI applications in the energy sector towards Agentic Digital TwinsLLM agents and tool use → Financial trading and forecasting agents
2ndMatch: Finetuning Pruned Diffusion Models via Second-Order Jacobian MatchingDiffusion models → Diffusion foundations and sampling
Leaps Beyond the Seen: Reinforced Reasoning Augmented Generation for Clinical NotesLLM reasoning → Clinical and biomedical applications
Q-Ponder: A Unified Training Pipeline for Reasoning-based Visual Quality AssessmentLLM reasoning → Multimodal reasoning +1
Understanding the Impact of Sampling Quality in Direct Preference OptimizationLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
Quake: Adaptive Indexing for Vector SearchRetrieval-augmented generation → Vector search and ANN indexes
DistRAG: Towards Distance-Based Spatial Reasoning in LLMsLLM reasoning → Spatial and embodied reasoning
The Impact of On-Policy Parallelized Data Collection on Deep Reinforcement Learning NetworksDeep reinforcement learning methods → Deep RL algorithms and representations
RAGOps: Operating and Managing Retrieval-Augmented Generation PipelinesRetrieval-augmented generation → RAG pipeline design and evaluation
SNIFR : Boosting Fine-Grained Child Harmful Content Detection Through Audio-Visual Alignment with Cascaded Cross-TransformerLLM safety and security → Vision-language and text-to-image safety
Toward Reliable VLM: A Fine-Grained Benchmark and Framework for Exposure, Bias, and Inference in Korean Street ViewsMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
Adversarial Attacks on Robotic Vision Language Action ModelsVision-language-action models → Safety and adversarial attacks +1
Seeing the Arrow of Time in Large Multimodal ModelsMultimodal LLMs / vision-language models → Video understanding with MLLMs
A Differential Perspective on Distributional Reinforcement LearningDeep reinforcement learning methods → RL theory and sample complexity
Helpful Agent Meets Deceptive Judge: Understanding Vulnerabilities in Agentic WorkflowsLLM agents and tool use → General LLM agent frameworks and benchmarks
APEX: Asynchronous Parallel CPU-GPU Execution for Online LLM Inference on Constrained GPUsEfficient LLM inference and serving → Serving systems and request scheduling
Unleashing the Reasoning Potential of Pre-trained LLMs by Critique Fine-Tuning on One ProblemLLM reasoning → Reinforcement learning for reasoning
HyperSteer: Activation Steering at Scale with HypernetworksMechanistic interpretability → Activation steering methods
Chipmunk: Training-Free Acceleration of Diffusion Transformers with Dynamic Column-Sparse DeltasDiffusion models → Diffusion foundations and sampling
BadReward: Clean-Label Poisoning of Reward Models in Text-to-Image RLHFLLM alignment / RLHF / preference optimisation → Diffusion and multimodal model alignment
Multiple-Frequencies Population-Based TrainingDeep reinforcement learning methods → Deep RL algorithms and representations
ConMamba: Contrastive Vision Mamba for Plant Disease DetectionState-space sequence models (S4, Mamba) → Vision Mamba backbones
GUI-Actor: Coordinate-Free Visual Grounding for GUI AgentsLLM agents and tool use → GUI and mobile agents
Not All Tokens Are Meant to Be ForgottenLLM safety and security → Machine unlearning
SVGenius: Benchmarking LLMs in SVG Understanding, Editing and GenerationMultimodal LLMs / vision-language models → Chart understanding and anomaly detection
OmniSpatial: Towards Comprehensive Spatial Reasoning Benchmark for Vision Language ModelsLLM reasoning → Spatial and embodied reasoning +1
Dual-Expert Consistency Model for Efficient and High-Quality Video GenerationDiffusion models → Video generation
Controllable Human-centric Keyframe Interpolation with Generative PriorDiffusion models → Video generation
Targeted Forgetting of Image Subgroups in CLIP ModelsLLM safety and security → Machine unlearning
Rectified Flows for Fast Multiscale Fluid Flow ModelingDiffusion models → Flow matching, time series and science
Critique-GRPO: Advancing LLM Reasoning with Natural Language and Numerical FeedbackLLM reasoning → Reinforcement learning for reasoning
Retrieval-Augmented Generation as Noisy In-Context Learning: A Unified Theory and Risk BoundsRetrieval-augmented generation → RAG pipeline design and evaluation
TalkingMachines: Real-Time Audio-Driven FaceTime-Style Video via Autoregressive Diffusion ModelsDiffusion models → Audio, speech and music
EgoVLM: Policy Optimization for Egocentric Video UnderstandingLLM reasoning → Spatial and embodied reasoning
DPO Learning with LLMs-Judge Signal for Computer Use AgentsMultimodal LLMs / vision-language models → Embodied and 3D vision-language models +2
From Flat to Hierarchical: Extracting Sparse Representations with Matching PursuitMechanistic interpretability → Sparse autoencoder training and evaluation
InterMamba: Efficient Human-Human Interaction Generation with Adaptive Spatio-Temporal MambaState-space sequence models (S4, Mamba) → Video, motion and generative Mamba models
ORV: 4D Occupancy-centric Robot Video GenerationWorld models → Embodied and robot world models
EDITOR: Effective and Interpretable Prompt Inversion for Text-to-Image Diffusion ModelsDiffusion models → Image generation, text-to-image and editing
Sign-SZPO: Provable Preference-based Reinforcement Learning with an Unknown Link FunctionDeep reinforcement learning methods → RL training of language and agent models
Sparse-vDiT: Unleashing the Power of Sparse Attention to Accelerate Video Diffusion TransformersDiffusion models → Video generation
MAEBE: Multi-Agent Emergent Behavior FrameworkLLM agents and tool use → Social simulation and multi-agent debate
EDEN: Entorhinal Driven Egocentric Navigation Toward Robotic DeploymentDeep reinforcement learning methods → RL for robot navigation and UAV control
Towards Analyzing and Understanding the Limitations of VAPO: A Theoretical PerspectiveLLM reasoning → Reinforcement learning for reasoning
InfiniteAudio: Infinite-Length Audio Generation with ConsistencyDiffusion models → Audio, speech and music
Coding Agents with Multimodal Browsing are Generalist Problem SolversLLM agents and tool use → Web, search, and research agents
DFBench: Benchmarking Deepfake Image Detection Capability of Large Multimodal ModelsMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness
PartComposer: Learning and Composing Part-Level Concepts from Single-Image ExamplesDiffusion models → Image generation, text-to-image and editing
A Multi-Agent Framework for Mitigating Dialect Biases in Privacy Policy Question-Answering SystemsLLM agents and tool use → Agent threats, guardrails and governance
Linear Spatial World Models Emerge in Large Language ModelsWorld models → World models for language agents
Mitigating Manipulation and Enhancing Persuasion: A Reflective Multi-Agent Approach for Legal Argument GenerationLLM agents and tool use → Social simulation and multi-agent debate
PC-MoE: Memory-Efficient and Privacy-Preserving Collaborative Training for Mixture-of-Experts LLMsMixture-of-experts → Expert routing strategies
UniConFlow: A Unified Constrained Flow-Matching Framework for Certified Motion PlanningDiffusion models → Diffusion and flow policies for robotics
Adaptive Graph Pruning for Multi-Agent CommunicationLLM agents and tool use → General LLM agent frameworks and benchmarks
Abstract Counterfactuals for Language Model AgentsLLM agents and tool use → General LLM agent frameworks and benchmarks
Hallucination to Consensus: Multi-Agent LLMs for End-to-End JUnit Test GenerationLLM agents and tool use → Software engineering coding agents
MTL-KD: Multi-Task Learning Via Knowledge Distillation for Generalizable Neural Vehicle Routing SolverKnowledge distillation → On-policy and self-distillation for LLMs
World Modelling Improves Language Model AgentsWorld models → World models for language agents +2
Language-Guided Generation for Personalized Inspection PlanningMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
Cell-o1: Training LLMs to Solve Single-Cell Reasoning Puzzles with Reinforcement LearningLLM reasoning → Clinical and biomedical applications
Diffusion Buffer: Online Diffusion-based Speech Enhancement with Sub-Second LatencyDiffusion models → Audio, speech and music
FlySearch: Exploring how vision-language models exploreMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
Scaling Fine-Grained MoE Beyond 50B Parameters: Empirical Evaluation and Practical InsightsMixture-of-experts → MoE large language models
A Continual Offline Reinforcement Learning Benchmark for Navigation TasksOffline RL and imitation learning from data → Decision transformers and sequence modeling
CoT is Not True Reasoning, It Is Just a Tight Constraint to Imitate: A Theory PerspectiveLLM reasoning → Prompting and chain-of-thought reasoning
It's the Thought that Counts: Evaluating the Attempts of Frontier LLMs to Persuade on Harmful TopicsLLM reasoning → Reasoning benchmarks and evaluation +1
Demystifying Reasoning Dynamics with Mutual Information: Thinking Tokens are Information Peaks in LLM ReasoningLLM reasoning → Test-time compute and efficient thinking
Surfer-H Meets Holo1: Cost-Efficient Web Agent Powered by Open WeightsMultimodal LLMs / vision-language models → Embodied and 3D vision-language models +1
BNPO: Beta Normalization Policy OptimizationLLM reasoning → Reinforcement learning for reasoning
ATAG: AI-Agent Application Threat Assessment with Attack GraphsLLM agents and tool use → Agent threats, guardrails and governance
DGMO: Training-Free Audio Source Separation through Diffusion-Guided Mask OptimizationDiffusion models → Audio, speech and music
METok: Multi-Stage Event-based Token Compression for Efficient Long Video UnderstandingEfficient LLM inference and serving → Visual token pruning in multimodal models
Learned Controllers for Agile Quadrotors in Pursuit-Evasion GamesDeep reinforcement learning methods → RL for robot navigation and UAV control
DeepShop: A Benchmark for Deep Research Shopping AgentsLLM agents and tool use → Social simulation and multi-agent debate
SemVink: Advancing VLMs' Semantic Understanding of Optical Illusions via Visual Global ThinkingLLM reasoning → Multimodal reasoning +1
Accelerating Model-Based Reinforcement Learning using Non-Linear Trajectory OptimizationDeep reinforcement learning methods → RL theory and sample complexity
Rethinking Machine Unlearning in Image Generation ModelsLLM safety and security → Machine unlearning
Why do AI agents communicate in human language?LLM agents and tool use → General LLM agent frameworks and benchmarks
Open-PMC-18M: A High-Fidelity Large Scale Medical Dataset for Multimodal Representation LearningMultimodal LLMs / vision-language models → Medical vision-language models
RACE-Align: Retrieval-Augmented and Chain-of-Thought Enhanced Preference Alignment for Large Language ModelsLLM reasoning → Reinforcement learning for reasoning +2
LocalGPT: Benchmarking and Advancing Large Language Models for Local Life Services in MeituanWorld models → World models for language agents
End-to-End Optimization of LLM-Driven Multi-Agent Search Systems via Heterogeneous-Group-Based Reinforcement LearningLLM agents and tool use → Reinforcement learning for LLM agents
Open-Set Living Need Prediction with Large Language ModelsWorld models → World models for language agents
Privacy Leaks by Adversaries: Adversarial Iterations for Membership Inference AttackLLM safety and security → Membership inference attacks
Iterative Self-Improvement of Vision Language Models for Image Scoring and Self-ExplanationMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning +1
On Entity Identification in Language ModelsMechanistic interpretability → Knowledge and representations inside LLMs
Smoothed Preference Optimization via ReNoise Inversion for Aligning Diffusion Models with Varied Human PreferencesDiffusion models → Diffusion and flow policies for robotics +1
LayoutRAG: Retrieval-Augmented Model for Content-agnostic Conditional Layout GenerationRetrieval-augmented generation → Multimodal RAG
Towards Geometry Problem Solving in the Large Model Era: A SurveyLLM reasoning → Spatial and embodied reasoning
Decompose, Plan in Parallel, and Merge: A Novel Paradigm for Large Language Models based Planning with Multiple ConstraintsLLM agents and tool use → General LLM agent frameworks and benchmarks
Solving Inverse Problems with FLAIRDiffusion models → Flow matching, time series and science
TL;DR: Too Long, Do Re-weighting for Efficient LLM Reasoning CompressionLLM reasoning → Test-time compute and efficient thinking
FAuNO: Semi-Asynchronous Federated Reinforcement Learning Framework for Task Offloading in Edge SystemsDeep reinforcement learning methods → RL for energy, networks and operations

All 507