Papers

New on arXiv

Mon 25 Aug 2025

346 papers

How can AI agents support journalists' work? An experiment with designing an LLM-driven intelligent reporting systemLLM agents and tool use → Web, search, and research agents
Multilevel Analysis of Cryptocurrency News using RAG Approach with Fine-Tuned Mistral Large Language ModelRetrieval-augmented generation → Legal and financial RAG
Robotic Fire Risk Detection based on Dynamic Knowledge Graph Reasoning: An LLM-Driven Approach with Graph Chain-of-ThoughtLLM reasoning → Spatial and embodied reasoning
Traj-MLLM: Can Multimodal Large Language Models Reform Trajectory Data Mining?Multimodal LLMs / vision-language models → MLLM training and visual instruction tuning
DemoBias: An Empirical Study to Trace Demographic Biases in Vision Foundation ModelsMultimodal LLMs / vision-language models → Social bias and fairness in VLMs
Object Detection with Multimodal Large Vision-Language Models: An In-depth ReviewMultimodal LLMs / vision-language models → Driving VLMs and open-vocabulary detection
Stand on The Shoulders of Giants: Building JailExpert from Previous Attack ExperienceLLM safety and security → Jailbreaking
Sparse Autoencoders for Low-N Protein Function Prediction and DesignMechanistic interpretability → Interpreting biomedical and science models
BTW: A Non-Parametric Variance Stabilization Framework for Multimodal Model IntegrationMixture-of-experts → Multimodal MoE models
Breaking Through Barren Plateaus: Reinforcement Learning Initializations for Deep Variational Quantum CircuitsDeep reinforcement learning methods → Quantum reinforcement learning
Language Models For Generalised PDDL Planning: Synthesising Sound and Programmatic PoliciesWorld models → World models for language agents
Data Augmentation Improves Machine UnlearningLLM safety and security → Machine unlearning
DRTA: Dynamic Reward Scaling for Reinforcement Learning in Time Series Anomaly DetectionDeep reinforcement learning methods → Deep RL algorithms and representations
How Reliable are LLMs for Reasoning on the Re-ranking task?LLM reasoning → Reasoning benchmarks and evaluation
CLARIFY: A Specialist-Generalist Framework for Accurate and Lightweight Dermatological Visual Question AnsweringMultimodal LLMs / vision-language models → Medical vision-language models
Toward Generalized Autonomous Agents: A Neuro-Symbolic AI Framework for Integrating Social and Technical Support in EducationLLM agents and tool use → Educational agents and math problem solving
Mining the Long Tail: A Comparative Study of Data-Centric Criticality Metrics for Robust Offline Reinforcement Learning in Autonomous Motion PlanningOffline RL and imitation learning from data → Imitation and offline RL for driving
Latent Self-Consistency for Reliable Majority-Set Selection in Short- and Long-Answer ReasoningLLM reasoning → Test-time compute and efficient thinking
PKG-DPO: Optimizing Domain-Specific AI systems with Physics Knowledge Graphs and Direct Preference OptimizationLLM alignment / RLHF / preference optimisation → Process reward models for reasoning
Backprompting: Leveraging Synthetic Production Data for Health Advice GuardrailsLLM safety and security → Safety alignment, refusal and guardrails
Language-Specific Layer Matters: Efficient Multilingual Enhancement for Large Vision-Language ModelsMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
DualSparse-MoE: Coordinating Tensor/Neuron-Level Sparsity with Expert Partition and ReconstructionMixture-of-experts → MoE large language models
Training Language Model Agents to Find Vulnerabilities with CTF-DojoLLM agents and tool use → Agent threats, guardrails and governance
FlowVLA: Visual Chain of Thought-based Motion Reasoning for Vision-Language-Action ModelsWorld models → Embodied and robot world models +2
InternVL3.5: Advancing Open-Source Multimodal Models in Versatility, Reasoning, and EfficiencyLLM reasoning → Multimodal reasoning
MMTok: Multimodal Coverage Maximization for Efficient Inference of VLMsMultimodal LLMs / vision-language models → Visual token pruning and efficiency
MIRAGE: Scaling Test-Time Inference with Parallel Graph-Retrieval-Augmented Reasoning ChainsLLM reasoning → Knowledge- and tool-augmented reasoning +1
Hermes 4 Technical ReportLLM reasoning → Mathematical reasoning
Follow My Hold: Hand-Object Interaction Reconstruction through Geometric GuidanceDiffusion models → Diffusion and flow policies for robotics
Better Language Model-Based Judging Reward Modeling through Scaling Comprehension BoundariesLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF
SEAM: Semantically Equivalent Across Modalities Benchmark for Vision-Language ModelsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Scene-Aware Vectorized Memory Multi-Agent Framework with Cross-Modal Differentiated Quantization VLMs for Visually Impaired AssistanceLLM agents and tool use → Multimodal and embodied agents
Improving End-to-End Training of Retrieval-Augmented Generation Models via Joint Stochastic ApproximationRetrieval-augmented generation → RAG pipeline design and evaluation
Mirroring Users: Towards Building Preference-aligned User Simulator with User Feedback in RecommendationLLM alignment / RLHF / preference optimisation → Preference alignment for recommenders and RAG
Test-Time Scaling Strategies for Generative Retrieval in Multimodal Conversational RecommendationsLLM reasoning → Knowledge- and tool-augmented reasoning
Provable Mixed-Noise Learning with Flow-MatchingDiffusion models → Flow matching, time series and science
Incorporating Pre-trained Diffusion Models in Solving the Schr"odinger Bridge ProblemDiffusion models → Diffusion foundations and sampling
Teaching LLMs to Think Mathematically: A Critical Study of Decision-Making via OptimizationLLM reasoning → Mathematical reasoning
ArgusCogito: Chain-of-Thought for Cross-Modal Synergy and Omnidirectional Reasoning in Camouflaged Object SegmentationLLM reasoning → Multimodal reasoning
PerPilot: Personalizing VLM-based Mobile Agents via Memory and ExplorationMultimodal LLMs / vision-language models → Embodied and 3D vision-language models +1
Visual-CoG: Stage-Aware Reinforcement Learning with Chain of Guidance for Text-to-Image GenerationDeep reinforcement learning methods → RL training of language and agent models
Neural Algorithmic Reasoners informed Large Language Model for Multi-Agent Path FindingLLM reasoning → Reasoning for agents, planning and code +1
A holistic perception system of internal and external monitoring for ground autonomous vehicles: AutoTRUST paradigmEnd-to-end autonomous driving → End-to-end system architecture
Language Models Coupled with Metacognition Can Outperform Reasoning ModelsLLM reasoning → Test-time compute and efficient thinking
Adaptive Learning for Moving Target defence: Enhancing Cybersecurity StrategiesDeep reinforcement learning methods → Adversarial robustness and RL in finance
Egocentric Instruction-oriented Affordance Prediction via Large Multimodal ModelMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
Pandora: Leveraging Code-driven Knowledge Transfer for Unified Structured Knowledge ReasoningLLM reasoning → Knowledge- and tool-augmented reasoning
ILRe: Intermediate Layer Retrieval for Context Compression in Causal Language ModelsEfficient LLM inference and serving → Prompt and context compression
ISALux: Illumination and Segmentation Aware Transformer Employing Mixture of Experts for Low Light Image EnhancementMixture-of-experts → Vision mixture-of-experts
FasterVoiceGrad: Faster One-step Diffusion-Based Voice Conversion with Adversarial Diffusion Conversion DistillationDiffusion models → Audio, speech and music
Retrieval Feedback Memory Enhancement Large Model Retrieval Generation MethodRetrieval-augmented generation → RAG pipeline design and evaluation
AVAM: Universal Training-free Adaptive Visual Anchoring Embedded into Multimodal Large Language Model for Multi-image Question AnsweringMultimodal LLMs / vision-language models → Visual token pruning and efficiency
LexSemBridge: Fine-Grained Dense Representation Enhancement through Token-Aware Embedding AugmentationRetrieval-augmented generation → Generation and context integration
Beyond Demographics: Enhancing Cultural Value Survey Simulation with Multi-Stage Personality-Driven Cognitive ReasoningLLM reasoning → Reasoning benchmarks and evaluation
GEPO: Group Expectation Policy Optimization for Stable Heterogeneous Reinforcement LearningDeep reinforcement learning methods → Deep RL algorithms and representations
Alternating Training-based Label Smoothing Enhances Prompt GeneralizationMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
Diffusion-Based Data Augmentation for Medical Image SegmentationDiffusion models → Medical imaging and segmentation
Multi-domain Distribution Learning for De Novo Drug DesignDiffusion models → Flow matching, time series and science
Attention Debiasing for Token Pruning in Vision Language ModelsMultimodal LLMs / vision-language models → Visual token pruning and efficiency +1
DRQA: Dynamic Reasoning Quota Allocation for Controlling Overthinking in Reasoning Large Language ModelsLLM reasoning → Test-time compute and efficient thinking
Proximal Supervised Fine-TuningDeep reinforcement learning methods → RL training of language and agent models
AgentRAN: An Agentic AI Architecture for Autonomous Control of Open 6G NetworksLLM agents and tool use → Agentic AI for wireless networks
CEIDM: A Controlled Entity and Interaction Diffusion Model for Enhanced Text-to-Image GenerationDiffusion models → Image generation, text-to-image and editing
SuperGen: An Efficient Ultra-high-resolution Video Generation System with Sketching and TilingDiffusion models → Video generation
Multi-layer Abstraction for Nested Generation of Options (MANGO) in Hierarchical Reinforcement LearningDeep reinforcement learning methods → Deep RL algorithms and representations
Speculative Safety-Aware DecodingEfficient LLM inference and serving → Speculative decoding +1
RepoTransAgent: Multi-Agent LLM Framework for Repository-Aware Code TranslationLLM agents and tool use → Software engineering coding agents
Instant Preference Alignment for Text-to-Image Diffusion ModelsDiffusion models → Diffusion and flow policies for robotics +1
F2RVLM: Boosting Fine-grained Fragment Retrieval for Multi-Modal Long-form Dialogue with Vision Language ModelMultimodal LLMs / vision-language models → Document understanding and multimodal RAG
EMPOWER: Evolutionary Medical Prompt Optimization With Reinforcement LearningDeep reinforcement learning methods → RL for clinical treatment and surgery
LLM-based Agentic Reasoning Frameworks: A Survey from Methods to ScenariosLLM reasoning → Reasoning for agents, planning and code +1
On the Edge of Memorization in Diffusion ModelsDiffusion models → Diffusion foundations and sampling
Language-Guided Temporal Token Pruning for Efficient VideoLLM ProcessingMultimodal LLMs / vision-language models → Visual token pruning and efficiency +1
Unlearning as Ablation: Toward a Falsifiable Benchmark for Generative Scientific DiscoveryLLM safety and security → Machine unlearning
Characterizing the Behavior of Training Mamba-based State Space Models on GPUsState-space sequence models (S4, Mamba) → Core SSM architectures for language
Towards Synthesizing Normative Data for Cognitive Assessments Using Generative Multimodal Large Language ModelsMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
Attacking LLMs and AI Agents: Advertisement Embedding Attacks Against Large Language ModelsLLM agents and tool use → Agent threats, guardrails and governance
Dynamic Embedding of Hierarchical Visual Features for Efficient Vision-Language Fine-TuningMultimodal LLMs / vision-language models → Visual token pruning and efficiency
Weights-Rotated Preference Optimization for Large Language ModelsLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
Finding Outliers in a Haystack: Anomaly Detection for Large Pointcloud ScenesState-space sequence models (S4, Mamba) → Point clouds, detection and remote sensing
ControlEchoSynth: Boosting Ejection Fraction Estimation Models via Controlled Video DiffusionDiffusion models → Medical imaging and segmentation
The Evolution of Thought: Tracking LLM Overthinking via Reasoning Dynamics AnalysisLLM reasoning → Test-time compute and efficient thinking
ExpertWeave: Efficiently Serving Expert-Specialized Fine-Tuned Adapters at ScaleMixture-of-experts → Mixture of LoRA experts
EMO-Reasoning: Benchmarking Emotional Reasoning Capabilities in Spoken Dialogue SystemsLLM reasoning → Audio-language model reasoning
Steering When Necessary: Flexible Steering Large Language Models with BacktrackingMechanistic interpretability → Activation steering methods
FlowRec: Prior-Informed Flow Matching for Efficient Sequential Recommendation GenerationDiffusion models → Diffusion recommenders
JCo-MVTON: Jointly Controllable Multi-Modal Diffusion Transformer for Mask-Free Virtual Try-onDiffusion models → 3D, point cloud and scene generation
GWM: Towards Scalable Gaussian World Models for Robotic ManipulationWorld models → Embodied and robot world models
TinyGiantVLM: A Lightweight Vision-Language Architecture for Spatial Reasoning under Resource ConstraintsLLM reasoning → Spatial and embodied reasoning +2
RubikSQL: Lifelong Learning Agentic Knowledge Base as an Industrial NL2SQL SystemLLM agents and tool use → Web, search, and research agents
HERO: Hierarchical Extrapolation and Refresh for Efficient World ModelsWorld models → Video-generation world models and simulators +1
Humanizing Machines: Rethinking LLM Anthropomorphism Through a Multi-Level Framework of DesignLLM reasoning → Reasoning benchmarks and evaluation
TradingGroup: A Multi-Agent Trading System with Self-Reflection and Data-SynthesisLLM agents and tool use → Financial trading and forecasting agents
How good are LLMs at Retrieving Documents in a Specific Domain?
DLGE: Dual Local-Global Encoding for Generalizable Cross-BCI-Paradigm
Signals vs. Videos: Advancing Motion Intention Recognition for Human-Robot Collaboration in Construction
Reading Between the Signs: Predicting Future Suicidal Ideation from Adolescent Social Media Texts
Multimodal Proposal for an AI-Based Tool to Increase Cross-Assessment of Messages
The ProLiFIC dataset: Leveraging LLMs to Unveil the Italian Lawmaking Process
Enhancing Speech Large Language Models through Reinforced Behavior Alignment

All 346