Papers

New on arXiv

Mon 14 Apr 2025

361 papers

Building Trustworthy Multimodal AI: A Review of Fairness, Transparency, and Ethics in Vision-Language TasksMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
LLM-based AI Agent for Sizing of Analog and Mixed Signal CircuitLLM agents and tool use → Software engineering coding agents
IEA-Plugin: An AI Agent Reasoner for Test Data AnalyticsLLM reasoning → Reasoning for agents, planning and code +1
Toward Aligning Human and Robot Actions via Multi-Modal Demonstration LearningOffline RL and imitation learning from data → Robot manipulation imitation learning
Uncovering Branch specialization in InceptionV1 using k sparse autoencodersMechanistic interpretability → Sparse autoencoder training and evaluation
How Instruction and Reasoning Data shape Post-Training: Data Quality through the Lens of Layer-wise GradientsLLM reasoning → Reasoning benchmarks and evaluation
ReasonDrive: Efficient Visual Question Answering for Autonomous Vehicles with Reasoning-Enhanced Small Vision-Language ModelsLLM reasoning → Chain-of-thought reasoning for driving +2
Foundation Models for Remote Sensing: An Analysis of MLLMs for Object LocalizationMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
HELIOS: Adaptive Model And Early-Exit Selection for Efficient LLM Inference ServingEfficient LLM inference and serving → Early exit and edge inference
SpinMeRound: Consistent Multi-View Identity Generation Using Diffusion ModelsDiffusion models → Audio, speech and music
The Jailbreak Tax: How Useful are Your Jailbreak Outputs?LLM safety and security → Jailbreaking
Relation-Rich Visual Document Generator for Visual Information ExtractionMultimodal LLMs / vision-language models → Document understanding and multimodal RAG
On the Contractivity of Stochastic Interpolation FlowDiffusion models → Diffusion foundations and sampling
Improving In-Context Learning with Reasoning DistillationLLM reasoning → Prompting and chain-of-thought reasoning +1
Weight-of-Thought Reasoning: Exploring Neural Network Weights for Enhanced LLM ReasoningLLM reasoning → Prompting and chain-of-thought reasoning
SilVar-Med: A Speech-Driven Visual Language Model for Explainable Abnormality Detection in Medical ImagingMultimodal LLMs / vision-language models → Medical vision-language models
Better Estimation of the Kullback--Leibler Divergence Between Language ModelsLLM alignment / RLHF / preference optimisation → LLM preference optimization methods +1
Beyond Chains of Thought: Benchmarking Latent-Space Reasoning Abilities in Large Language ModelsLLM reasoning → Reasoning benchmarks and evaluation
Energy Matching: Unifying Flow Matching and Energy-Based Models for Generative ModelingDiffusion models → Flow matching, time series and science
Demo: ViolentUTF as An Accessible Platform for Generative AI Red TeamingLLM safety and security → Prompt injection and agent security
AgMMU: A Comprehensive Agricultural Multimodal Understanding BenchmarkMultimodal LLMs / vision-language models → Agricultural vision-language models
H3AE: High Compression, High Speed, and High Quality AutoEncoder for Video Diffusion ModelsDiffusion models → Diffusion foundations and sampling
Efficient Process Reward Model Training via Active LearningLLM reasoning → Reinforcement learning for reasoning +1
REPA-E: Unlocking VAE for End-to-End Tuning with Latent Diffusion TransformersDiffusion models → Diffusion foundations and sampling
xVerify: Efficient Answer Verifier for Reasoning Model EvaluationsLLM reasoning → Reasoning benchmarks and evaluation
Weight Ensembling Improves Reasoning in Language ModelsLLM reasoning → Test-time compute and efficient thinking
Pixel-SAIL: Single Transformer For Pixel-Grounded UnderstandingMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
The Scalability of Simplicity: Empirical Analysis of Vision-Language Learning with a Single TransformerMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
GUI-R1 : A Generalist R1-Style Vision-Language Action Model For GUI AgentsMultimodal LLMs / vision-language models → Embodied and 3D vision-language models +2
M1: Towards Scalable Test-Time Compute with Mamba Reasoning ModelsLLM reasoning → Test-time compute and efficient thinking +1
Anchor Token Matching: Implicit Structure Locking for Training-free AR Image EditingDiffusion models → Image generation, text-to-image and editing
MonoDiff9D: Monocular Category-Level 9D Object Pose Estimation via Diffusion ModelDiffusion models → 3D, point cloud and scene generation
SymRTLO: Enhancing RTL Code Optimization with LLMs and Neuron-Inspired Symbolic ReasoningLLM reasoning → Reasoning for agents, planning and code
Exploring the System 1 Thinking Capability of Large Reasoning ModelsLLM reasoning → Test-time compute and efficient thinking
FingER: Content Aware Fine-grained Evaluation with Reasoning for AI-Generated VideosLLM reasoning → Multimodal reasoning
VisualPuzzles: Decoupling Multimodal Reasoning Evaluation from Domain KnowledgeLLM reasoning → Multimodal reasoning
Heimdall: test-time scaling on the generative verificationLLM reasoning → Mathematical reasoning
InstructEngine: Instruction-driven Text-to-Image AlignmentLLM alignment / RLHF / preference optimisation → Diffusion and multimodal model alignment
AlayaDB: The Data Foundation for Efficient and Effective Long-context LLM InferenceEfficient LLM inference and serving → Serving systems and request scheduling
SlowFastVAD: Video Anomaly Detection via Integrating Simple Detector and RAG-Enhanced Vision-Language ModelMultimodal LLMs / vision-language models → Chart understanding and anomaly detection
Analysis of Attention in Video Diffusion TransformersDiffusion models → Video generation
AutoStyle-TTS: Retrieval-Augmented Generation based Automatic Style Matching Text-to-Speech SynthesisRetrieval-augmented generation → Conversational and speech RAG
Characterizing LLM-driven Social Network: The Chirper.ai CaseLLM agents and tool use → Social simulation and multi-agent debate
α-Flow: A Unified Framework for Continuous-State Discrete Flow Matching ModelsDiffusion models → Flow matching, time series and science
Vision based driving agent for race car simulation environmentsEnd-to-end autonomous driving → Reinforcement learning for driving +1
XY-Cut++: Advanced Layout Ordering via Hierarchical Mask Mechanism on a Novel BenchmarkRetrieval-augmented generation → Multimodal RAG
Adaptive Sensor Steering Strategy Using Deep Reinforcement Learning for Dynamic Data Acquisition in Digital TwinsDeep reinforcement learning methods → RL for robot locomotion and manipulation
PRM-BAS: Enhancing Multimodal Reasoning through PRM-guided Beam Annealing SearchMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Can Competition Enhance the Proficiency of Agents Powered by Large Language Models in the Realm of News-driven Time Series Forecasting?LLM agents and tool use → Financial trading and forecasting agents
From Prompting to Alignment: A Generative Framework for Query RecommendationLLM alignment / RLHF / preference optimisation → Preference alignment for recommenders and RAG
DioR: Adaptive Cognitive Detection and Contextual Retrieval Optimization for Dynamic Retrieval-Augmented GenerationRetrieval-augmented generation → RAG pipeline design and evaluation
Localized Cultural Knowledge is Conserved and Controllable in Large Language ModelsWorld models → World models for language agents
DeepTrans: Deep Reasoning Translation via Reinforcement LearningLLM reasoning → Reasoning benchmarks and evaluation
LLM Unlearning Reveals a Stronger-Than-Expected Coreset Effect in Current BenchmarksLLM safety and security → Machine unlearning
The Future of MLLM Prompting is Adaptive: A Comprehensive Experimental Evaluation of Prompt Engineering Methods for Robust Multimodal PerformanceMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
MSCoT: Structured Chain-of-Thought Generation for Multiple Programming LanguagesLLM reasoning → Reasoning for agents, planning and code
HalluSearch at SemEval-2025 Task 3: A Search-Enhanced RAG Pipeline for Hallucination DetectionRetrieval-augmented generation → Hallucination detection in RAG
MT-R1-Zero: Advancing LLM-based Machine Translation via R1-Zero-like Reinforcement LearningLLM reasoning → Reasoning benchmarks and evaluation
COUNTS: Benchmarking Object Detectors and Multimodal Large Language Models under Distribution ShiftsMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
SocioVerse: A World Model for Social Simulation Powered by LLM Agents and A Pool of 10 Million Real-World UsersWorld models → World models for language agents +1
Hierarchical and Step-Layer-Wise Tuning of Attention Specialty for Multi-Instance Synthesis in Diffusion TransformersDiffusion models → Image generation, text-to-image and editing
A Survey of Personalization: From RAG to AgentRetrieval-augmented generation → Agentic RAG, agents and memory +1
Breaking the Data Barrier -- Building GUI Agents Through Task GeneralizationLLM agents and tool use → GUI and mobile agents
Global and Local Mamba Network for Multi-Modality Medical Image Super-ResolutionState-space sequence models (S4, Mamba) → Mamba for medical image analysis
A Human-Sensitive Controller: Adapting to Human Musculoskeletal Disorder-Related Constraints via Reinforcement LearningDeep reinforcement learning methods → RL for robot locomotion and manipulation
CameraBench: Benchmarking Visual Reasoning in MLLMs via PhotographyLLM reasoning → Multimodal reasoning +1
RealSafe-R1: Safety-Aligned DeepSeek-R1 without Compromising Reasoning CapabilityLLM reasoning → Reasoning for agents, planning and code +1
Towards Quantifying Commonsense Reasoning with Mechanistic InsightsLLM reasoning → Reasoning benchmarks and evaluation
MMKB-RAG: A Multi-Modal Knowledge-Based Retrieval-Augmented Generation FrameworkMultimodal LLMs / vision-language models → Document understanding and multimodal RAG +1
Pay Attention to What and Where? Interpretable Feature Extractor in Vision-based Deep Reinforcement LearningDeep reinforcement learning methods → Interpretable and logic-guided RL
Mavors: Multi-granularity Video Representation for Multimodal Large Language ModelMultimodal LLMs / vision-language models → Video understanding with MLLMs
Summarization of Multimodal Presentations with Vision-Language Models: Study of the Effect of Modalities and StructureMultimodal LLMs / vision-language models → Video understanding with MLLMs
GraphCodeAgent: Dual Graph-Guided LLM Agent for Retrieval-Augmented Repo-Level Code GenerationRetrieval-augmented generation → RAG for code +1
CHARM: Calibrating Reward Models With Chatbot Arena ScoresLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF
Aligning Anime Video Generation with Human FeedbackLLM alignment / RLHF / preference optimisation → Diffusion and multimodal model alignment
Prior Does Matter: Visual Navigation via Denoising Diffusion Bridge ModelsDiffusion models → Diffusion and flow policies for robotics +1
The Mirage of Performance Gains: Why Contrastive Decoding Fails to Mitigate Object Hallucinations in MLLMs?Multimodal LLMs / vision-language models → Hallucination mitigation in LVLMs
Improving Controller Generalization with Dimensionless Markov Decision ProcessesDeep reinforcement learning methods → RL theory and sample complexity
NaviDiffusor: Cost-Guided Diffusion Model for Visual NavigationDiffusion models → Diffusion and flow policies for robotics
FLoRA: Sample-Efficient Preference-based RL via Low-Rank Style Adaptation of Reward FunctionsLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF +1
GaussVideoDreamer: 3D Scene Generation with Video Diffusion and Inconsistency-Aware Gaussian SplattingDiffusion models → 3D, point cloud and scene generation
Do We Really Need Curated Malicious Data for Safety Alignment in Multi-modal Large Language Models?Multimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness +2
Correlative and Discriminative Label Grouping for Multi-Label Visual Prompt TuningMixture-of-experts → Multimodal MoE models
Resampling Benchmark for Efficient Comprehensive Evaluation of Large Vision-Language ModelsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
OctGPT: Octree-based Multiscale Autoregressive Models for 3D Shape GenerationDiffusion models → 3D, point cloud and scene generation
Assessing Judging Bias in Large Reasoning Models: An Empirical StudyLLM reasoning → Reasoning benchmarks and evaluation
KeepKV: Achieving Periodic Lossless KV Cache Compression for Efficient LLM InferenceEfficient LLM inference and serving → KV cache optimization
Moderate Actor-Critic Methods: Controlling Overestimation Bias via Expectile LossDeep reinforcement learning methods → RL theory and sample complexity
Efficient Task-specific Conditional Diffusion Policies: Shortcut Model Acceleration and SO(3) OptimizationDiffusion models → Diffusion and flow policies for robotics +1
FLARE: Fully Integration of Vision-Language Representations for Deep Cross-Modal UnderstandingMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
Guiding Reasoning in Small Language Models with LLM AssistanceLLM reasoning → Reasoning benchmarks and evaluation
Learning to Erase Private Knowledge from Multi-Documents for Retrieval-Augmented Large Language ModelsRetrieval-augmented generation → Privacy and security of RAG
Plasticity-Aware Mixture of Experts for Learning Under QoE Shifts in Adaptive Video StreamingMixture-of-experts → Mixture of LoRA experts
Digital Staining with Knowledge Distillation: A Unified Framework for Unpaired and Paired-But-Misaligned DataKnowledge distillation → Distillation for medical imaging
TAMP: Token-Adaptive Layerwise Pruning in Multimodal Large Language ModelsMultimodal LLMs / vision-language models → Visual token pruning and efficiency
Learning from Reference Answers: Versatile Language Model Alignment without Binary Human Preference DataLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
Separate to Collaborate: Dual-Stream Diffusion Model for Coordinated Piano Hand Motion SynthesisDiffusion models → Audio, speech and music
Reasoning Models Can Be Effective Without ThinkingLLM reasoning → Test-time compute and efficient thinking
PestMA: LLM-based Multi-Agent System for Informed Pest ManagementLLM agents and tool use → General LLM agent frameworks and benchmarks
A Survey of Large Language Model-Powered Spatial Intelligence Across Scales: Advances in Embodied Agents, Smart Cities, and Earth ScienceWorld models → World models for language agents +1

All 361