Papers

New on arXiv

Tue 8 Jul 2025

361 papers

AgentMaster: A Multi-Agent Conversational Framework Using A2A and MCP Protocols for Multimodal Information Retrieval and AnalysisLLM agents and tool use → General LLM agent frameworks and benchmarks
Assuring the Safety of Reinforcement Learning Components: AMLAS-RLDeep reinforcement learning methods → Safe and constrained RL
A Systematic Analysis of Declining Medical Safety Messaging in Generative AI ModelsMultimodal LLMs / vision-language models → Medical vision-language models
Self-Consistency in Vision-Language Models for Precision Agriculture: Multi-Response Consensus for Crop Disease ManagementLLM reasoning → Clinical and biomedical applications +1
Circumventing Safety Alignment in Large Language Models Through Embedding Space Toxicity AttenuationLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks +1
Automated Neuron Labelling Enables Generative Steering and Interpretability in Protein Language ModelsMechanistic interpretability → Interpreting biomedical and science models
FedPhD: Federated Pruning with Hierarchical Learning of Diffusion ModelsDiffusion models → Robustness, privacy and adversarial aspects
Perception-Aware Policy Optimization for Multimodal ReasoningLLM reasoning → Multimodal reasoning
Neural Actor-Critic Methods for Hamilton-Jacobi-Bellman PDEs: Asymptotic Analysis and Numerical StudiesDeep reinforcement learning methods → RL theory and sample complexity
Teach a Reward Model to Correct Itself: Reward Guided Adversarial Failure Discovery for Robust Reward ModelingLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF
PERK: Long-Context Reasoning as Test-Time LearningLLM reasoning → Test-time compute and efficient thinking
Learning to Evaluate Autonomous Behaviour in Human-Robot InteractionOffline RL and imitation learning from data → Robot manipulation imitation learning
Representing Prompting Patterns with PDL: Compliance Agent Case StudyLLM agents and tool use → General LLM agent frameworks and benchmarks
Mitigating 3D Prostate Biparametric MRI Data Scarcity through Domain Adaptation using Locally-Trained Latent Diffusion Models for Prostate Cancer DetectionDiffusion models → Medical imaging and segmentation
Mamba Goes HoME: Hierarchical Soft Mixture-of-Experts for 3D Medical Image SegmentationMixture-of-experts → Medical and biomedical MoE +1
Sample-Efficient Reinforcement Learning Controller for Deep Brain Stimulation in Parkinson's DiseaseDeep reinforcement learning methods → Deep RL algorithms and representations
Bridging AI and Software Security: A Comparative Vulnerability Assessment of LLM Agent Deployment ParadigmsLLM agents and tool use → Agent threats, guardrails and governance
Too Human to Model:The Uncanny Valley of LLMs in Social Simulation -- When Generative Language Agents Misalign with Modelling PrinciplesLLM agents and tool use → Social simulation and multi-agent debate
The bitter lesson of misuse detectionLLM safety and security → Jailbreaking
Enhancing LLM Watermark Resilience Against Both Scrubbing and Spoofing AttacksLLM safety and security → Watermarking generated content and models
LIRA: Inferring Segmentation in Large Multi-modal Models with Local Interleaved Region AssistanceMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Agent KB: Leveraging Cross-Domain Experience for Agentic Problem SolvingLLM agents and tool use → Agent memory systems
EC-Flow: Enabling Versatile Robotic Manipulation from Action-Unlabeled Videos via Embodiment-Centric FlowOffline RL and imitation learning from data → Robot manipulation imitation learning
Aligned Textual Scoring RulesLLM alignment / RLHF / preference optimisation → Human values and AI feedback alignment
CultureCLIP: Empowering CLIP with Cultural Awareness through Synthetic Images and Contextualized CaptionsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Differential MambaState-space sequence models (S4, Mamba) → Core SSM architectures for language
A Survey on Latent ReasoningLLM reasoning → Test-time compute and efficient thinking
Hidden Prompts in Manuscripts Exploit AI-Assisted Peer ReviewLLM safety and security → Prompt injection and agent security
Enhancing Scientific Visual Question Answering through Multimodal Reasoning and Ensemble ModelingLLM reasoning → Multimodal reasoning
CriticLean: Critic-Guided Reinforcement Learning for Mathematical FormalizationDeep reinforcement learning methods → RL training of language and agent models
Learning Agile Tensile Perching for Aerial Robots from DemonstrationsDeep reinforcement learning methods → RL for robot locomotion and manipulation
Skywork-R1V3 Technical ReportLLM reasoning → Multimodal reasoning +1
Prompt-Free Conditional Diffusion for Multi-object Image AugmentationDiffusion models → Image generation, text-to-image and editing
LangMamba: A Language-driven Mamba Framework for Low-dose CT Denoising with Vision-language ModelsMultimodal LLMs / vision-language models → Medical vision-language models +1
OpenAgentSafety: A Comprehensive Framework for Evaluating Real-World AI Agent SafetyLLM agents and tool use → Agent threats, guardrails and governance
Brownian Bridge Diffusion for Sequential RecommendationDiffusion models → Diffusion recommenders
Speech Quality Assessment Model Based on Mixture of Experts: System-Level Performance Enhancement and Utterance-Level Challenge AnalysisMixture-of-experts → Multimodal MoE models
Safe Domain Randomization via Uncertainty-Aware Out-of-Distribution Detection and Policy AdaptationDeep reinforcement learning methods → Safe and constrained RL
CoRE: Enhancing Metacognition with Label-free Self-evaluation in LRMsLLM reasoning → Test-time compute and efficient thinking
QS4D: Quantization-aware training for efficient hardware deployment of structured state-space sequential modelsState-space sequence models (S4, Mamba) → Core SSM architectures for language
ScoreAdv: Score-based Targeted Generation of Natural Adversarial Examples via Diffusion ModelsDiffusion models → Robustness, privacy and adversarial aspects
MCAM: Multimodal Causal Analysis Model for Ego-Vehicle-Level Driving Video UnderstandingEnd-to-end autonomous driving → End-to-end system architecture
FEVO: Financial Knowledge Expansion and Reasoning Evolution for Large Language ModelsLLM reasoning → Reasoning for agents, planning and code
CAVGAN: Unifying Jailbreak and Defense of LLMs via Generative Adversarial Attacks on their Internal RepresentationsLLM safety and security → Jailbreaking
TextPixs: Glyph-Conditioned Diffusion with Character-Aware Attention and OCR-Guided SupervisionDiffusion models → Image generation, text-to-image and editing
Conditional Multi-Stage Failure Recovery for Embodied AgentsLLM agents and tool use → Software engineering coding agents
CogniSQL-R1-Zero: Lightweight Reinforced Reasoning for Efficient SQL GenerationLLM reasoning → Knowledge- and tool-augmented reasoning
Multi-Agent Debate Strategies to Enhance Requirements Engineering with Large Language ModelsLLM agents and tool use → Software engineering coding agents
Lost in Localization: Building RabakBench with Human-in-the-Loop Validation to Measure Multilingual Safety GapsLLM safety and security → Safety alignment, refusal and guardrails
T-LoRA: Single Image Diffusion Model Customization Without OverfittingDiffusion models → Image generation, text-to-image and editing
Tora2: Motion and Appearance Customized Diffusion Transformer for Multi-Entity Video GenerationDiffusion models → Video generation
BlueLM-2.5-3B Technical ReportLLM reasoning → Multimodal reasoning +1
High-Resolution Visual Reasoning via Multi-Turn Grounding-Based Reinforcement LearningLLM reasoning → Multimodal reasoning +1
Accelerating Diffusion Model Training under Minimal Budgets: A Condensation-Based PerspectiveDiffusion models → Diffusion foundations and sampling
Best-of-N through the Smoothing Lens: KL Divergence and Regret AnalysisLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
Differentiable Reward Optimization for LLM based TTS systemLLM alignment / RLHF / preference optimisation → Preference alignment for recommenders and RAG
Feature-Based vs. GAN-Based Learning from Demonstrations: When and WhyOffline RL and imitation learning from data → Imitation learning methods
Stable Acoustic Relay Assignment with High Throughput via Lase Chaos-based Reinforcement LearningDeep reinforcement learning methods → RL for robot navigation and UAV control
GeoMag: A Vision-Language Model for Pixel-level Fine-Grained Remote Sensing Image ParsingMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
Current Practices for Building LLM-Powered Reasoning Tools Are Ad Hoc -- and We Can Do BetterLLM reasoning → Reasoning benchmarks and evaluation
KERAG_R: Knowledge-Enhanced Retrieval-Augmented Generation for RecommendationRetrieval-augmented generation → RAG for recommendation
Video Event Reasoning and Prediction by Fusing World Knowledge from LLMs with Vision Foundation ModelsLLM reasoning → Multimodal reasoning
Constella: Supporting Storywriters' Interconnected Character Creation through LLM-based Multi-AgentsLLM agents and tool use → Multimodal and embodied agents
Concept-Based Mechanistic Interpretability Using Structured Knowledge GraphsMechanistic interpretability → Circuits and mechanistic analysis
Bridging Perception and Language: A Systematic Benchmark for LVLMs' Understanding of Amodal Completion ReportsMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
SPADE: Spatial-Aware Denoising Network for Open-vocabulary Panoptic Scene Graph Generation with Long- and Local-range Context ReasoningLLM reasoning → Spatial and embodied reasoning +1
GTA1: GUI Test-time Scaling AgentLLM reasoning → Multimodal reasoning
TalkFashion: Intelligent Virtual Try-On Assistant Based on Multimodal Large Language ModelMultimodal LLMs / vision-language models → Image editing and quality assessment
Robust Bandwidth Estimation for Real-Time Communication with Offline Reinforcement LearningOffline RL and imitation learning from data → Offline multi-agent RL
LeAD: The LLM Enhanced Planning System Converged with End-to-end Autonomous DrivingEnd-to-end autonomous driving → Large language models for driving
Non-Intrusive Binaural Speech Intelligibility Prediction Using Mamba for Hearing-Impaired ListenersState-space sequence models (S4, Mamba) → Speech and audio SSMs
Omni-Router: Sharing Routing Decisions in Sparse Mixture-of-Experts for Speech RecognitionMixture-of-experts → Multimodal MoE models
Large Language Models for Agent-Based Modelling: Current and possible uses across the modelling cycleLLM agents and tool use → General LLM agent frameworks and benchmarks
Hierarchical Task Offloading for UAV-Assisted Vehicular Edge Computing via Deep Reinforcement LearningDeep reinforcement learning methods → RL for energy, networks and operations
MobileGUI-RL: Advancing Mobile GUI Agent through Reinforcement Learning in Online EnvironmentLLM agents and tool use → GUI and mobile agents +1
HIRAG: Hierarchical-Thought Instruction-Tuning Retrieval-Augmented GenerationLLM reasoning → Knowledge- and tool-augmented reasoning +1
DRAGOn: Designing RAG On Periodically Updated CorpusRetrieval-augmented generation → RAG pipeline design and evaluation
Agentic-R1: Distilled Dual-Strategy ReasoningLLM reasoning → Mathematical reasoning +1
Hybrid Diffusion Policies with Projective Geometric Algebra for Efficient Robot Manipulation LearningDiffusion models → Diffusion and flow policies for robotics
Robust One-step Speech Enhancement via Consistency DistillationDiffusion models → Audio, speech and music
Efficient Training of Large-Scale AI Models Through Federated Mixture-of-Experts: A System-Level ApproachMixture-of-experts → Federated mixture-of-experts
LiON-LoRA: Rethinking LoRA Fusion to Unify Controllable Spatial and Temporal Generation for Video DiffusionDiffusion models → Video generation
Integrated Structural Prompt Learning for Vision-Language ModelsMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
Integrating Diffusion-based Multi-task Learning with Online Reinforcement Learning for Robust Quadruped Robot ControlDiffusion models → Diffusion and flow policies for robotics +2
R-VLM: Region-Aware Vision Language Model for Precise GUI GroundingMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
Modeling and Reversing Brain Lesions Using Diffusion ModelsDiffusion models → Medical imaging and segmentation
Dynamic Rank Adaptation for Vision-Language ModelsMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
Knowledge-guided Complex Diffusion Model for PolSAR Image Classification in Contourlet DomainDiffusion models → Inverse problems and image restoration
Optimus: A Robust Defense Framework for Mitigating Toxicity while Fine-Tuning Conversational AILLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks +1
HRRRCast: a data-driven emulator for regional weather forecasting at convection allowing scalesDiffusion models → Flow matching, time series and science
ECom-Bench: Can LLM Agent Resolve Real-World E-commerce Customer Support Issues?LLM agents and tool use → Social simulation and multi-agent debate
SARA: Selective and Adaptive Retrieval-augmented Generation with Context CompressionRetrieval-augmented generation → Generation and context integration +1
How Not to Detect Prompt Injections with an LLMLLM safety and security → Prompt injection and agent security
ADMC: Attention-based Diffusion Model for Missing Modalities Feature CompletionDiffusion models → Image generation, text-to-image and editing
Detecting Proxy Gaming in RL and LLM Alignment via Evaluator Stress TestsLLM alignment / RLHF / preference optimisation → Reward hacking and misspecification
Flipping Knowledge Distillation: Leveraging Small Models' Expertise to Enhance LLMs in Text MatchingKnowledge distillation → Knowledge distillation for LLMs
Structured Task Solving via Modular Embodied Intelligence: A Case Study on Rubik's CubeMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
Kernel Density Steering: Inference-Time Scaling via Mode Seeking for Image RestorationDiffusion models → Inverse problems and image restoration
Model-free Optical Processors using In Situ Reinforcement Learning with Proximal Policy OptimizationDeep reinforcement learning methods → RL for energy, networks and operations
The Landscape of Memorization in LLMs: Mechanisms, Measurement, and MitigationLLM safety and security → Machine unlearning

All 361