Papers

New on arXiv

Thu 10 Apr 2025

305 papers

MALIBU Benchmark: Multi-Agent LLM Implicit Bias UncoveredLLM agents and tool use → Social simulation and multi-agent debate
Marmot: Object-Level Self-Correction via Multi-Agent ReasoningLLM reasoning → Multimodal reasoning +2
Seed1.5-Thinking: Advancing Superb Reasoning Models with Reinforcement LearningLLM reasoning → Reasoning benchmarks and evaluation +1
AttentionDefense: Leveraging System Prompt Attention for Explainable Defense Against Novel JailbreaksLLM safety and security → Jailbreaking
Data Metabolism: An Efficient Data Design Schema For Vision Language ModelMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Capybara-OMNI: An Efficient Paradigm for Building Omni-Modal Language ModelsMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
SFT or RL? An Early Investigation into Training R1-Like Reasoning Large Vision-Language ModelsLLM reasoning → Multimodal reasoning +1
ColorBench: Can VLMs See and Understand the Colorful World? A Comprehensive Benchmark for Color Perception, Reasoning, and RobustnessLLM reasoning → Multimodal reasoning
SD2: Self-Distilled Sparse DraftersEfficient LLM inference and serving → Speculative decoding
VL-Rethinker: Incentivizing Self-Reflection of Vision-Language Models with Reinforcement LearningLLM reasoning → Multimodal reasoning +1
Rethinking the Foundations for Continual Reinforcement LearningDeep reinforcement learning methods → Deep RL algorithms and representations
Investigating Vision-Language Model for Point Cloud-based Vehicle ClassificationMultimodal LLMs / vision-language models → Driving VLMs and open-vocabulary detection
Orchestrating Agents and Data for Enterprise: A Blueprint Architecture for Compound AILLM agents and tool use → General LLM agent frameworks and benchmarks
DeepSeek-R1 vs. o3-mini: How Well can Reasoning LLMs Evaluate MT and Summarization?LLM reasoning → Reasoning benchmarks and evaluation
RL-based Control of UAS Subject to Significant DisturbanceDeep reinforcement learning methods → RL for robot navigation and UAV control
Test Amplification for REST APIs via Single and Multi-Agent LLM SystemsLLM agents and tool use → Software engineering coding agents
POEM: Precise Object-level Editing via MLLM controlDiffusion models → Image generation, text-to-image and editing
Geneshift: Impact of different scenario shift on Jailbreaking LLMLLM safety and security → Jailbreaking
ContrastiveGaussian: High-Fidelity 3D Generation with Contrastive Learning and Gaussian SplattingDiffusion models → 3D, point cloud and scene generation
Deep Reinforcement Learning for Day-to-day Dynamic Tolling in Tradable Credit SchemesDeep reinforcement learning methods → Adversarial robustness and RL in finance
The AI Scientist-v2: Workshop-Level Automated Scientific Discovery via Agentic Tree SearchLLM agents and tool use → Web, search, and research agents
Compositional Flows for 3D Molecule and Synthesis Pathway Co-designDiffusion models → Flow matching, time series and science
Can Reasoning LLMs Enhance Clinical Document Classification?LLM reasoning → Clinical and biomedical applications
DGFamba: Learning Flow Factorized State Space for Visual Domain GeneralizationState-space sequence models (S4, Mamba) → Vision Mamba backbones
Cat, Rat, Meow: On the Alignment of Language Model and Human Term-Similarity JudgmentsLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
C3PO: Critical-Layer, Core-Expert, Collaborative Pathway Optimization for Test-Time Expert Re-MixingMixture-of-experts → MoE large language models
GLUS: Global-Local Reasoning Unified into A Single Large Language Model for Video SegmentationLLM reasoning → Multimodal reasoning +1
Geo4D: Leveraging Video Generators for Geometric 4D Scene ReconstructionDiffusion models → 3D, point cloud and scene generation
VisualCloze: A Universal Image Generation Framework via Visual In-Context LearningDiffusion models → Image generation, text-to-image and editing
MM-IFEngine: Towards Multimodal Instruction FollowingMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning +1
VCR-Bench: A Comprehensive Evaluation Framework for Video Chain-of-Thought ReasoningLLM reasoning → Multimodal reasoning +1
Perception-R1: Pioneering Perception Policy with Reinforcement LearningLLM reasoning → Multimodal reasoning +1
GenEAva: Generating Cartoon Avatars with Fine-Grained Facial Expressions from Realistic Diffusion-based FacesDiffusion models → Audio, speech and music
Echo: An Open-Source, Low-Cost Teleoperation System with Force Feedback for Dataset Collection in Robot LearningOffline RL and imitation learning from data → Robot manipulation imitation learning
SoTA with Less: MCTS-Guided Sample Selection for Data-Efficient Visual Reasoning Self-ImprovementLLM reasoning → Multimodal reasoning +2
Echo Chamber: RL Post-training Amplifies Behaviors Learned in PretrainingLLM reasoning → Reinforcement learning for reasoning
How do Large Language Models Understand Relevance? A Mechanistic Interpretability PerspectiveMechanistic interpretability → Knowledge and representations inside LLMs
Fast Adaptation with Behavioral Foundation ModelsDeep reinforcement learning methods → Deep RL algorithms and representations
SpecReason: Fast and Accurate Inference-Time Compute via Speculative ReasoningLLM reasoning → Test-time compute and efficient thinking +1
Benchmarking Adversarial Robustness to Bias Elicitation in Large Language Models: Scalable Automated Assessment with LLM-as-a-JudgeLLM safety and security → Safety alignment, refusal and guardrails
Dual-Difficulty Curriculum Learning for Direct Preference OptimizationLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
Deceptive Automated Interpretability: Language Models Coordinating to Fool Oversight SystemsMechanistic interpretability → Probing deception, bias and beliefs
MOSAIC: Modeling Social AI for Content Dissemination and Regulation in Multi-Agent SimulationsLLM agents and tool use → Social simulation and multi-agent debate
Who Benchmarks the Benchmarks? Towards Comprehensive Evaluation of Commonsense Reasoning BenchmarksLLM reasoning → Reasoning benchmarks and evaluation
Cluster-Driven Expert Pruning for Mixture-of-Experts Large Language ModelsMixture-of-experts → MoE large language models
A System for Comprehensive Assessment of RAG FrameworksRetrieval-augmented generation → RAG pipeline design and evaluation
Plan-and-Refine: Diverse and Comprehensive Retrieval-Augmented GenerationRetrieval-augmented generation → RAG robustness and adaptive retrieval
Genetic Programming with Reinforcement Learning Trained Transformer for Real-World Dynamic Scheduling ProblemsDeep reinforcement learning methods → RL for energy, networks and operations
SlimSpeech: Lightweight and Efficient Text-to-Speech with Slim Rectified FlowDiffusion models → Audio, speech and music
Virtual-mask Informed Prior for Sparse-view Dual-Energy CT ReconstructionDiffusion models → Medical imaging and segmentation
SF2T: Self-supervised Fragment Finetuning of Video-LLMs for Fine-Grained UnderstandingMultimodal LLMs / vision-language models → Video understanding with MLLMs
The Multi-Round Diagnostic RAG Framework for Emulating Clinical ReasoningLLM reasoning → Clinical and biomedical applications +1
A Relative Ignorability Framework for Decision-Relevant Observability in Control Theory and Reinforcement LearningDeep reinforcement learning methods → RL theory and sample complexity
PR-Attack: Coordinated Prompt-RAG Attacks on Retrieval-Augmented Generation in Large Language Models via Bilevel OptimizationRetrieval-augmented generation → Privacy and security of RAG
Sim-to-Real Transfer in Reinforcement Learning for Maneuver Control of a Variable-Pitch MAVDeep reinforcement learning methods → RL for robot navigation and UAV control
Distilling Knowledge from Heterogeneous Architectures for Semantic SegmentationKnowledge distillation → Detection and dense prediction distillation
ms-Mamba: Multi-scale Mamba for Time-Series ForecastingState-space sequence models (S4, Mamba) → Time series forecasting with SSMs
On the Temporal Question-Answering Capabilities of Large Language Models Over Anonymized DataLLM reasoning → Knowledge- and tool-augmented reasoning
CollEX -- A Multimodal Agentic RAG System Enabling Interactive Exploration of Scientific CollectionsRetrieval-augmented generation → Multimodal RAG +1
Enhancing Large Language Models through Neuro-Symbolic Integration and Ontological ReasoningLLM reasoning → Reasoning benchmarks and evaluation
Agent That Debugs: Dynamic State-Guided Vulnerability RepairLLM agents and tool use → Software engineering coding agents
ConceptFormer: Towards Efficient Use of Knowledge-Graph Embeddings in Large Language ModelsRetrieval-augmented generation → Knowledge-graph and structured-knowledge RAG
VLM-R1: A Stable and Generalizable R1-style Large Vision-Language ModelLLM reasoning → Multimodal reasoning +1
Boosting Universal LLM Reward Design through Heuristic Reward Observation Space EvolutionLLM reasoning → Reinforcement learning for reasoning
REANIMATOR: Reanimate Retrieval Test Collections with Extracted and Synthetic ResourcesRetrieval-augmented generation → Generation and context integration
Exploring Human-Like Thinking in Search Simulations with Large Language ModelsLLM reasoning → Reasoning benchmarks and evaluation
Diffusion Transformers for Tabular Data Time Series GenerationDiffusion models → Flow matching, time series and science
STeP: A Framework for Solving Scientific Video Inverse Problems with Spatiotemporal Diffusion PriorsDiffusion models → Inverse problems and image restoration
AI-Slop to AI-Polish? Aligning Language Models through Edit-Based Writing Rewards and Test-time ComputationLLM alignment / RLHF / preference optimisation → Human values and AI feedback alignment
Why We Feel: Breaking Boundaries in Emotional Reasoning with Multimodal Large Language ModelsLLM reasoning → Audio-language model reasoning +1
Drive in Corridors: Enhancing the Safety of End-to-end Autonomous Driving via Corridor Learning and PlanningEnd-to-end autonomous driving → End-to-end system architecture
Apt-Serve: Adaptive Request Scheduling on Hybrid Cache for Scalable LLM Inference ServingEfficient LLM inference and serving → Serving systems and request scheduling
Kimi-VL Technical ReportLLM reasoning → Multimodal reasoning +2
UniCAIM: A Unified CAM/CIM Architecture with Static-Dynamic KV Cache Pruning for Efficient Long-Context LLM InferenceEfficient LLM inference and serving → KV cache optimization
Defense against Prompt Injection Attacks via Mixture of EncodingsLLM safety and security → Prompt injection and agent security
Achilles Heel of Distributed Multi-Agent SystemsLLM agents and tool use → Agent threats, guardrails and governance
How Can Objects Help Video-Language Understanding?Multimodal LLMs / vision-language models → MLLM training and visual instruction tuning
Unifying and extending Diffusion Models through PDEs for solving Inverse ProblemsDiffusion models → Inverse problems and image restoration
Enhancing Player Enjoyment with a Two-Tier DRL and LLM-Based Agent System for Fighting GamesLLM agents and tool use → Reinforcement learning for LLM agents
Routing to the Right Expertise: A Trustworthy Judge for Instruction-based Image EditingMultimodal LLMs / vision-language models → Image editing and quality assessment +1
ThermoStereoRT: Thermal Stereo Matching in Real Time via Knowledge Distillation and Attention-based RefinementKnowledge distillation → Detection and dense prediction distillation
RA-RRG: Multimodal Retrieval-Augmented Radiology Report Generation with Key Phrase ExtractionRetrieval-augmented generation → Medical and biomedical RAG
A Novel Mamba-based Sequential Recommendation MethodState-space sequence models (S4, Mamba) → Core SSM architectures for language
State Estimation Using Particle Filtering in Adaptive Machine Learning Methods: Integrating Q-Learning and NEAT Algorithms with Noisy Radar MeasurementsDeep reinforcement learning methods → RL theory and sample complexity
ID-Booth: Identity-consistent Face Generation with Diffusion ModelsDiffusion models → Image generation, text-to-image and editing
PROPEL: Supervised and Reinforcement Learning for Large-Scale Supply Chain PlanningDeep reinforcement learning methods → RL for energy, networks and operations
Novel Diffusion Models for Multimodal 3D Hand Trajectory PredictionDiffusion models → Diffusion and flow policies for robotics
Revisiting Prompt Optimization with Large Reasoning Models-A Case Study on Event ExtractionLLM reasoning → Prompting and chain-of-thought reasoning
Throughput-Optimal Scheduling Algorithms for LLM Inference and AI AgentsLLM agents and tool use → Software engineering coding agents +1
Evaluating LLMs for Visualization Tasks
APSQ: Additive Partial Sum Quantization with Algorithm-Hardware Co-Design
ZS-VCOS: Zero-Shot Video Camouflaged Object Segmentation By Optical Flow and Open Vocabulary Object Detection
A constraints-based approach to fully interpretable neural networks for detecting learner behaviors
Can Geometry Save Central Views for Sports Field Registration?
Evaluating Large Language Models on Multiword Expressions in Multilingual and Code-Switched Contexts
Parameter Tuning of the Firefly Algorithm by Three Tuning Methods: Standard Monte Carlo, Quasi-Monte Carlo and Latin Hypercube Sampling Methods
Critical Challenges and Guidelines in Evaluating Synthetic Tabular Data: A Systematic Review
My Precious Crash Data: Barriers and Opportunities in Encouraging Autonomous Driving Companies to Share Safety-Critical Data
S2Vec: Self-Supervised Geospatial Embeddings for the Built Environment
Memory-efficient Streaming VideoLLMs for Real-time Procedural Video Understanding

All 305