Papers

New on arXiv

Mon 7 Apr 2025

340 papers

A Desideratum for Conversational Agents: Capabilities, Challenges, and Future DirectionsLLM agents and tool use → General LLM agent frameworks and benchmarks
Social Human Robot Embodied Conversation (SHREC) Dataset: Benchmarking Foundational Models' Social ReasoningLLM reasoning → Spatial and embodied reasoning
CCSK:Cognitive Convection of Self-Knowledge Based Retrieval Augmentation for Large Language ModelsRetrieval-augmented generation → RAG robustness and adaptive retrieval
Prima.cpp: Fast 30-70B LLM Inference on Heterogeneous and Low-Resource Home ClustersEfficient LLM inference and serving → Speculative decoding
SEAL: Steerable Reasoning Calibration of Large Language Models for FreeLLM reasoning → Test-time compute and efficient thinking +1
Well2Flow: Reconstruction of reservoir states from sparse wells using score-based generative modelsDiffusion models → Flow matching, time series and science
Ternarization of Vision Language Models for use on edge devicesMultimodal LLMs / vision-language models → Visual token pruning and efficiency
Federated Hierarchical Reinforcement Learning for Adaptive Traffic Signal ControlDeep reinforcement learning methods → RL for autonomous driving and traffic
COIG-P: A High-Quality and Large-Scale Chinese Preference Dataset for Alignment with Human ValuesLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
Bridging Industrial Expertise and XR with LLM-Powered Conversational AgentsRetrieval-augmented generation → RAG pipeline design and evaluation +1
Deep Reinforcement Learning Algorithms for Option HedgingDeep reinforcement learning methods → Adversarial robustness and RL in finance
Efficient Reinforcement Finetuning via Adaptive Curriculum LearningLLM reasoning → Reinforcement learning for reasoning
Evaluating the Generalization Capabilities of Large Language Models on Code ReasoningLLM reasoning → Reasoning for agents, planning and code
REEF: Relevance-Aware and Efficient LLM Adapter for Video UnderstandingMultimodal LLMs / vision-language models → Video understanding with MLLMs
GraphRAFT: Retrieval Augmented Fine-Tuning for Knowledge Graphs on Graph DatabasesLLM reasoning → Knowledge- and tool-augmented reasoning +1
Trust Through Transparency: Explainable Social Navigation for Autonomous Mobile Robots via Vision-Language ModelsMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
Studying Image Diffusion Features for Zero-Shot Video Object SegmentationDiffusion models → Medical imaging and segmentation
Taxonomy-Aware Evaluation of Vision-Language ModelsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
ViewBridge: Curriculum Knowledge Distillation for Activity View-Invariance Under Extreme Viewpoint ChangesKnowledge distillation → Detection and dense prediction distillation
Probing the Visualization Literacy of Vision Language Models: the Good, the Bad, and the UglyMultimodal LLMs / vision-language models → Chart understanding and anomaly detection
EP-Diffuser: An Efficient Diffusion Model for Traffic Scene Generation and Prediction via Polynomial RepresentationsDiffusion models → Diffusion and flow policies for robotics
Reasoning Models Know When They're Right: Probing Hidden States for Self-VerificationLLM reasoning → Test-time compute and efficient thinking
RouteFormer: A Transformer-Based Routing Framework for Autonomous VehiclesEnd-to-end autonomous driving → Cooperative and V2X driving
The Role of Environment Access in Agnostic Reinforcement LearningDeep reinforcement learning methods → RL theory and sample complexity
Physics-informed Modularized Neural Network for Advanced Building Control by Deep Reinforcement LearningDeep reinforcement learning methods → RL for energy, networks and operations
Interactive Explanations for Reinforcement-Learning AgentsDeep reinforcement learning methods → Interpretable and logic-guided RL
EduPlanner: LLM-Based Multi-Agent Systems for Customized and Intelligent Instructional DesignLLM agents and tool use → Educational agents and math problem solving
Debate-Feedback: A Multi-Agent Framework for Efficient Legal Judgment PredictionLLM agents and tool use → Social simulation and multi-agent debate
Gaussian Mixture Flow Matching ModelsDiffusion models → Flow matching, time series and science
Dimension-Free Convergence of Diffusion Models for Approximate Gaussian MixturesDiffusion models → Diffusion foundations and sampling
SmolVLM: Redefining small and efficient multimodal modelsMultimodal LLMs / vision-language models → Visual token pruning and efficiency
One-Minute Video Generation with Test-Time TrainingState-space sequence models (S4, Mamba) → Video, motion and generative Mamba models
Truthful or Fabricated? Using Causal Attribution to Mitigate Reward Hacking in ExplanationsLLM reasoning → Reinforcement learning for reasoning +1
Enhancing LLM-Based Short Answer Grading with Retrieval-Augmented GenerationRetrieval-augmented generation → RAG robustness and adaptive retrieval
How to evaluate control measures for LLM agents? A trajectory from today to superintelligenceLLM agents and tool use → Agent threats, guardrails and governance +1
Learning to Reason Over Time: Timeline Self-Reflection for Improved Temporal Reasoning in Language ModelsLLM reasoning → Knowledge- and tool-augmented reasoning
A Reality Check of Vision-Language Pre-training in Radiology: Have We Progressed Using Text?Multimodal LLMs / vision-language models → Medical vision-language models
Vision-Language Model Predictive Control for Manipulation Planning and Trajectory GenerationMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
Reinforced Multi-teacher Knowledge Distillation for Efficient General Image Forgery Detection and LocalizationKnowledge distillation → Detection and dense prediction distillation
Utility-Focused LLM Annotation for Retrieval and Retrieval-Augmented GenerationRetrieval-augmented generation → RAG pipeline design and evaluation
Concise Reasoning via Reinforcement LearningLLM reasoning → Test-time compute and efficient thinking
The 1st Solution for 4th PVUW MeViS Challenge: Unleashing the Potential of Large Multimodal Models for Referring Video SegmentationMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
RLBayes: a Bayesian Network Structure Learning Algorithm via Reinforcement Learning-Based Search StrategyDeep reinforcement learning methods → RL theory and sample complexity
Evaluating Knowledge Graph Based Retrieval Augmented Generation Methods under Knowledge IncompletenessRetrieval-augmented generation → Knowledge-graph and structured-knowledge RAG
DDPM Score Matching and Distribution LearningDiffusion models → Diffusion foundations and sampling
CARE: Multilingual Human Preference Learning for Cultural AwarenessLLM alignment / RLHF / preference optimisation → Human values and AI feedback alignment
A Reinforcement Learning Method for Environments with Stochastic Variables: Post-Decision Proximal Policy Optimization with Dual Critic NetworksDeep reinforcement learning methods → RL theory and sample complexity
EffOWT: Transfer Visual Language Models to Open-World Tracking Efficiently and EffectivelyMultimodal LLMs / vision-language models → Video understanding with MLLMs
DA2Diff: Exploring Degradation-aware Adaptive Diffusion Priors for All-in-One Weather RestorationDiffusion models → Inverse problems and image restoration
VAPO: Efficient and Reliable Reinforcement Learning for Advanced Reasoning TasksLLM reasoning → Reinforcement learning for reasoning
AI for Climate Finance: Agentic Retrieval and Multi-Step Reasoning for Early Warning System InvestmentsLLM reasoning → Reasoning for agents, planning and code +2
State Tuning: State-based Test-Time Scaling on RWKV-7LLM reasoning → Test-time compute and efficient thinking
The Curse of CoT: On the Limitations of Chain-of-Thought in In-Context LearningLLM reasoning → Prompting and chain-of-thought reasoning
Not All Data Are Unlearned EquallyLLM safety and security → Machine unlearning
Revealing the Intrinsic Ethical Vulnerability of Aligned Large Language ModelsLLM alignment / RLHF / preference optimisation → Human values and AI feedback alignment
Debate Only When Necessary: Adaptive Multiagent Collaboration for Efficient LLM ReasoningLLM reasoning → Reasoning for agents, planning and code
Spatiotemporal Attention-Augmented Inverse Reinforcement Learning for Multi-Agent Task AllocationOffline RL and imitation learning from data → Inverse reinforcement learning
Graph-based Diffusion Model for Collaborative FilteringDiffusion models → Diffusion recommenders
Joint Pedestrian and Vehicle Traffic Optimization in Urban Environments using Reinforcement LearningDeep reinforcement learning methods → RL for autonomous driving and traffic
Wavelet Policy: Imitation Learning in the Scale Domain with World Prior MemoryOffline RL and imitation learning from data → Robot manipulation imitation learning
Remote Sensing Retrieval-Augmented Generation: Bridging Remote Sensing Imagery and Comprehensive Knowledge with a Multi-Modal Dataset and Retrieval-Augmented Generation ModelRetrieval-augmented generation → Multimodal RAG
A Domain-Based Taxonomy of Jailbreak Vulnerabilities in Large Language ModelsLLM safety and security → Jailbreaking
Towards Visual Text Grounding of Multimodal Large Language ModelMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Ensuring Safety in an Uncertain Environment: Constrained MDPs via Stochastic ThresholdsDeep reinforcement learning methods → RL theory and sample complexity
REWIND: Real-Time Egocentric Whole-Body Motion Diffusion with Exemplar-Based Identity ConditioningDiffusion models → Diffusion and flow policies for robotics
GOTHAM: Graph Class Incremental Learning Framework under Weak SupervisionKnowledge distillation → Graph neural network distillation
A Unified Pairwise Framework for RLHF: Bridging Generative Reward Modeling and Policy OptimizationLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF
A Llama walks into the 'Bar': Efficient Supervised Fine-Tuning for Legal Reasoning in the Multi-state Bar ExamLLM reasoning → Reasoning for agents, planning and code
Age-of-information minimization under energy harvesting and non-stationary environmentDeep reinforcement learning methods → RL for energy, networks and operations
Collab-RAG: Boosting Retrieval-Augmented Generation for Complex Question Answering via White-Box and Black-Box LLM CollaborationLLM reasoning → Knowledge- and tool-augmented reasoning +1
Video-Bench: Human-Aligned Video Generation BenchmarkLLM alignment / RLHF / preference optimisation → Diffusion and multimodal model alignment
SCAM: A Real-World Typographic Robustness Evaluation for Multimodal Foundation ModelsMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness
Simulating Persuasive Dialogues on Meat Reduction with Generative AgentsLLM agents and tool use → Social simulation and multi-agent debate
GAMDTP: Dynamic Trajectory Prediction with Graph Attention Mamba NetworkState-space sequence models (S4, Mamba) → Video, motion and generative Mamba models
BIASINSPECTOR: Detecting Bias in Structured Data through LLM AgentsLLM agents and tool use → Social simulation and multi-agent debate
Quantization Hurts Reasoning? An Empirical Study on Quantized Reasoning ModelsLLM reasoning → Test-time compute and efficient thinking
SUEDE:Shared Unified Experts for Physical-Digital Face Attack Detection EnhancementMixture-of-experts → Vision mixture-of-experts
SEEM: Exploiting Black-Box Text Attacks to Manipulate Tool SelectionLLM agents and tool use → Tool learning and function calling
ELT-Bench: An End-to-End Benchmark for Evaluating AI Agents on ELT PipelinesLLM agents and tool use → Web, search, and research agents
OrderChain: Towards General Instruct-Tuning for Stimulating the Ordinal Understanding Ability of MLLMMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
TabRep: Training Tabular Diffusion Models with a Simple and Effective Continuous RepresentationDiffusion models → Flow matching, time series and science
Dynamic Vision MambaState-space sequence models (S4, Mamba) → Vision Mamba backbones +1
Weak-for-Strong: Training Weak Meta-Agent to Harness Strong ExecutorsLLM agents and tool use → Software engineering coding agents
Disentangling Instruction Influence in Diffusion Transformers for Parallel Multi-Instruction-Guided Image EditingDiffusion models → Image generation, text-to-image and editing
Playing Non-Embedded Card-Based Games with Reinforcement LearningDeep reinforcement learning methods → RL for games and combinatorial optimization
OCC-MLLM-CoT-Alpha: Towards Multi-stage Occlusion Recognition Based on Large Language Models via 3D-Aware Supervision and Chain-of-Thoughts GuidanceLLM reasoning → Multimodal reasoning +1
Feedback-Enhanced Hallucination-Resistant Vision-Language Model for Real-Time Scene UnderstandingMultimodal LLMs / vision-language models → Hallucination mitigation in LVLMs
Improving Multilingual Retrieval-Augmented Language Models through Dialectic Reasoning ArgumentationsLLM reasoning → Knowledge- and tool-augmented reasoning +1
Unsupervised Estimation of Nonlinear Audio Effects: Comparing Diffusion-Based and Adversarial approachesDiffusion models → Audio, speech and music
SCRAMBLe : Enhancing Multimodal LLM Compositionality with Synthetic Preference DataMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
Synthetic Data Generation & Multi-Step RL for Reasoning & Tool UseLLM reasoning → Reinforcement learning for reasoning +2
T1: Tool-integrated Verification for Test-time Compute Scaling in Small Language ModelsLLM reasoning → Test-time compute and efficient thinking
LagKV: Lag-Relative Information of the KV Cache Tells Which Tokens Are ImportantEfficient LLM inference and serving → KV cache optimization
Provable Failure of Language Models in Learning Majority Boolean Logic via Gradient DescentLLM reasoning → Reasoning benchmarks and evaluation
R2Vul: Learning to Reason about Software Vulnerabilities with Reinforcement Learning and Structured Reasoning DistillationLLM reasoning → Reasoning for agents, planning and code +2
HypRL: Reinforcement Learning of Control Policies for HyperpropertiesDeep reinforcement learning methods → Deep RL algorithms and representations
ACE-RLHF: Automated Code Evaluation and Socratic Feedback Generation Tool using Large Language Models and Reinforcement Learning with Human FeedbackLLM alignment / RLHF / preference optimisation → Process reward models for reasoning
LEO-MINI: An Efficient Multimodal Large Language Model using Conditional Token Reduction and Mixture of Multi-Modal ExpertsMultimodal LLMs / vision-language models → Visual token pruning and efficiency +1
Autono: A ReAct-Based Highly Robust Autonomous Agent FrameworkLLM agents and tool use → General LLM agent frameworks and benchmarks
Machine Learning Solutions Integrated in an IoT Healthcare Platform for Heart Failure Risk Stratification

All 340