Papers

New on arXiv

Thu 2 Apr 2026

462 papers

QoS-Aware Token Scheduling and Private Data Valuation for Multi-Modal Agentic NetworksLLM agents and tool use → Agent threats, guardrails and governance
Domain Adaptation of Large Language Models for Polymer-Composite Additive Manufacturing Using Retrieval-Augmented Generation and Fine-TuningRetrieval-augmented generation → Medical and biomedical RAG
Mitigating Cross-Lingual Cultural Inconsistencies in LLMs via Consensus-Driven Preference OptimisationLLM alignment / RLHF / preference optimisation → Human values and AI feedback alignment
Agentopic: A Generative AI Agent Workflow for Explainable Topic ModelingLLM reasoning → Reasoning for agents, planning and code +1
Anchored Confabulation: Partial Evidence Non-Monotonically Amplifies Confident Hallucination in LLMsLLM reasoning → Hallucination detection and mitigation
Adaptive Memory Crystallization for Autonomous AI Agent Learning in Dynamic EnvironmentsDeep reinforcement learning methods → Deep RL algorithms and representations
A Comparative Theoretical Analysis of Entropy Control Methods in Reinforcement LearningLLM reasoning → Reinforcement learning for reasoning +1
AViS-Mamba: Adaptive Visual Steering of Audio State-Space Dynamics for Violence DetectionState-space sequence models (S4, Mamba) → Video, motion and generative Mamba models
Safety-Aligned 3D Object Detection: Single-Vehicle, Cooperative, and End-to-End PerspectivesEnd-to-end autonomous driving → VLMs for autonomous driving
Mitigating LLM biases toward spurious social contexts using direct preference optimizationLLM alignment / RLHF / preference optimisation → Human values and AI feedback alignment
VoxelCodeBench: Benchmarking 3D World Modeling Through Code GenerationWorld models → Video-generation world models and simulators +1
Understanding the Effects of Safety Unalignment on Large Language ModelsLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks +1
WSVD: Weighted Low-Rank Approximation for Fast and Efficient Execution of Low-Precision Vision-Language ModelsMultimodal LLMs / vision-language models → Visual token pruning and efficiency
Dependency-Guided Parallel Decoding in Discrete Diffusion Language ModelsDiffusion models → Discrete and text diffusion +1
Principled and Scalable Diversity-Aware Retrieval via Cardinality-Constrained Binary Quadratic ProgrammingRetrieval-augmented generation → RAG robustness and adaptive retrieval
Beyond Resolution Rates: Behavioral Drivers of Coding Agent Success and FailureLLM reasoning → Reasoning for agents, planning and code +1
Developer Experience with AI Coding Agents: HTTP Behavioral Signatures in Documentation PortalsLLM agents and tool use → Software engineering coding agents
Overconfidence and Calibration in Medical VQA: Empirical Findings and Hallucination-Aware MitigationMultimodal LLMs / vision-language models → Medical vision-language models
Interpretable Deep Reinforcement Learning for Element-level Bridge Life-cycle OptimizationDeep reinforcement learning methods → RL for energy, networks and operations
Tune to Learn: How Controller Gains Shape Robot Policy LearningDeep reinforcement learning methods → RL for robot locomotion and manipulation
Reinforcement Learning from Human Feedback: A Statistical PerspectiveLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF +1
An Explainable Vision-Language Model Framework with Adaptive PID-Tversky Loss for Lumbar Spinal Stenosis DiagnosisMultimodal LLMs / vision-language models → Medical vision-language models
I must delete the evidence: AI Agents Explicitly Cover up Fraud and Violent CrimeLLM agents and tool use → Agent threats, guardrails and governance
Token-Efficient Multimodal Reasoning via Image Prompt PackagingLLM reasoning → Multimodal reasoning +1
VLMs Need Words: Vision Language Models Ignore Visual Detail In Favor of Semantic AnchorsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Failing to Falsify: Evaluating and Mitigating Confirmation Bias in Language ModelsLLM reasoning → Reasoning benchmarks and evaluation
AIVV: Neuro-Symbolic LLM Agent-Integrated Verification and Validation for Trustworthy Autonomous SystemsLLM agents and tool use → Agent threats, guardrails and governance
Single-Agent LLMs Outperform Multi-Agent Systems on Multi-Hop Reasoning Under Equal Thinking Token BudgetsLLM reasoning → Reasoning for agents, planning and code +1
Do We Need Frontier Models to Verify Mathematical Proofs?LLM reasoning → Mathematical reasoning
Mitigating Data Scarcity in Spaceflight Applications for Offline Reinforcement Learning Using Physics-Informed Deep Generative ModelsOffline RL and imitation learning from data → Offline RL algorithms
Compositional Neuro-Symbolic ReasoningLLM reasoning → Multimodal reasoning
Generative models on phase spaceDiffusion models → Flow matching, time series and science
Beyond Fixed Inference: Quantitative Flow Matching for Adaptive Image DenoisingDiffusion models → Inverse problems and image restoration
ActionParty: Multi-Subject Action Binding in Generative Video GamesWorld models → Video-generation world models and simulators +1
Steerable Visual RepresentationsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Batched Contextual Reinforcement: A Task-Scaling Law for Efficient ReasoningLLM reasoning → Test-time compute and efficient thinking
Stop Wandering: Efficient Vision-Language Navigation via Metacognitive ReasoningLLM reasoning → Spatial and embodied reasoning
Omni123: Exploring 3D Native Foundation Models with Limited 3D Data by Unifying Text to 2D and 3D GenerationMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
Unifying Group-Relative and Self-Distillation Policy Optimization via Sample RoutingLLM reasoning → Reinforcement learning for reasoning +1
The Self Driving Portfolio: Agentic Architecture for Institutional Asset ManagementLLM agents and tool use → Financial trading and forecasting agents
SKILL0: In-Context Agentic Reinforcement Learning for Skill InternalizationLLM agents and tool use → Agent skill learning and evolution
Model-Based Reinforcement Learning for Control under Time-Varying DynamicsDeep reinforcement learning methods → RL theory and sample complexity
Retrieval-Augmented Question Answering over Scientific Literature for the Electron-Ion ColliderRetrieval-augmented generation → RAG robustness and adaptive retrieval
Smoothing the Landscape: Causal Structure Learning via Diffusion Denoising ObjectivesDiffusion models → Diffusion foundations and sampling
UAV-Track VLA: Embodied Aerial Tracking via Vision-Language-Action ModelsVision-language-action models → Aerial and drone VLA
Answering the Wrong Question: Reasoning Trace Inversion for Abstention in LLMsLLM reasoning → Reasoning benchmarks and evaluation
When to ASK: Uncertainty-Gated Language Assistance for Reinforcement LearningDeep reinforcement learning methods → Safe and constrained RL
Impact of Multimodal and Conversational AI on Learning Outcomes and ExperienceMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
Multi-Agent Video Recommenders: Evolution, Patterns, and Open ChallengesLLM agents and tool use → Agentic recommender systems
Where Does Robustness Live? Neuron-Guided Adaptation for Retrieval-Augmented Language ModelsRetrieval-augmented generation → RAG robustness and adaptive retrieval
UniDriveVLA: Unifying Understanding, Perception, and Action Planning for Autonomous DrivingVision-language-action models → VLA for autonomous driving +1
TRU: Targeted Reverse Update for Efficient Multimodal Recommendation UnlearningLLM safety and security → Membership inference attacks
The Expert Strikes Back: Interpreting Mixture-of-Experts Language Models at Expert LevelMixture-of-experts → MoE large language models
Quantifying Self-Preservation Bias in Large Language ModelsLLM alignment / RLHF / preference optimisation → Human values and AI feedback alignment
Reflection Generation for Composite Image Using Diffusion ModelDiffusion models → 3D, point cloud and scene generation
Brief Is Better: Non-Monotonic Chain-of-Thought Budget Effects in Function-Calling Language AgentsLLM reasoning → Prompting and chain-of-thought reasoning +1
Auction-Based Online Policy Adaptation for Evolving ObjectivesDeep reinforcement learning methods → RL theory and sample complexity
MTI: A Behavior-Based Temperament Profiling System for AI AgentsLLM agents and tool use → Social simulation and multi-agent debate
Reliable Control-Point Selection for Steering Reasoning in Large Language ModelsLLM reasoning → Test-time compute and efficient thinking +1
FlatAttention: Dataflow and Fabric Collectives Co-Optimization for Large Attention-Based Model Inference on Tile-Based AcceleratorsMixture-of-experts → Efficient MoE training and inference systems
GroundVTS: Visual Token Sampling in Multimodal Large Language Models for Video Temporal GroundingMultimodal LLMs / vision-language models → Video understanding with MLLMs
Optimizing RAG Rerankers with LLM Feedback via Reinforcement LearningRetrieval-augmented generation → Generation and context integration +1
PLUME: Latent Reasoning Based Universal Multimodal EmbeddingLLM reasoning → Multimodal reasoning
Jagle: Building a Large-Scale Japanese Multimodal Post-Training Dataset for Vision-Language ModelsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Goose: Anisotropic Speculation Trees for Training-Free Speculative DecodingEfficient LLM inference and serving → Speculative decoding
Efficient Reasoning via Thought Compression for Language SegmentationLLM reasoning → Test-time compute and efficient thinking +1
Reinforcement Learning for Speculative Trading under Exploratory FrameworkDeep reinforcement learning methods → Adversarial robustness and RL in finance
Why Gaussian Diffusion Models Fail on Discrete Data and How to Prevent It?Diffusion models → Diffusion foundations and sampling
Systematic Analyses of Reinforcement Learning Controllers in Signalized Urban CorridorsDeep reinforcement learning methods → RL for autonomous driving and traffic
ATBench: A Diverse and Realistic Agent Trajectory Benchmark for Safety Evaluation and DiagnosisLLM agents and tool use → Agent threats, guardrails and governance
Bridging Discrete Planning and Continuous Execution for Redundant RobotDeep reinforcement learning methods → RL for robot locomotion and manipulation
Apriel-Reasoner: RL Post-Training for General-Purpose and Efficient ReasoningLLM reasoning → Reinforcement learning for reasoning
ProCeedRL: Process Critic with Exploratory Demonstration Reinforcement Learning for LLM Agentic ReasoningLLM reasoning → Reinforcement learning for reasoning +1
Resonance4D: Frequency-Domain Motion Supervision for Preset-Free Physical Parameter Learning in 4D Dynamic Physical Scene SimulationDiffusion models → Video generation
SAFE: An LLM-as-Verifier Framework for Evidence-Grounded Multi-Hop ReasoningLLM reasoning → Knowledge- and tool-augmented reasoning
Attention at Rest Stays at Rest: Breaking Visual Inertia for Cognitive Hallucination MitigationMultimodal LLMs / vision-language models → Hallucination mitigation in LVLMs
World Action Verifier: Self-Improving World Models via Forward-Inverse AsymmetryWorld models → Embodied and robot world models
Ego-Grounding for Personalized Question-Answering in Egocentric VideosMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
The Rank and Gradient Lost in Non-stationarity: Sample Weight Decay for Mitigating Plasticity Loss in Reinforcement LearningDeep reinforcement learning methods → Deep RL algorithms and representations
Bayesian Elicitation with LLMs: Model Size Helps, Extra "Reasoning" Doesn't AlwaysLLM reasoning → Test-time compute and efficient thinking
Low-Effort Jailbreak Attacks Against Text-to-Image Safety FiltersLLM safety and security → Vision-language and text-to-image safety
When does learning pay off? A study on DRL-based dynamic algorithm configuration for carbon-aware schedulingDeep reinforcement learning methods → RL for energy, networks and operations
A3R: Agentic Affordance Reasoning via Cross-Dimensional Evidence in 3D Gaussian ScenesLLM reasoning → Spatial and embodied reasoning +1
HieraVid: Hierarchical Token Pruning for Fast Video Large Language ModelsEfficient LLM inference and serving → Visual token pruning in multimodal models
Towards Intrinsically Calibrated Uncertainty Quantification in Industrial Data-Driven Models via Diffusion SamplerDiffusion models → Flow matching, time series and science
Semantic Richness or Geometric Reasoning? The Fragility of VLM's Visual InvarianceLLM reasoning → Spatial and embodied reasoning +1
Not All Tokens See Equally: Perception-Grounded Policy Optimization for Large Vision-Language ModelsLLM reasoning → Multimodal reasoning +1
PLOT: Enhancing Preference Learning via Optimal TransportLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
Physics Informed Reinforcement Learning with Gibbs Priors for Topology Control in Power GridsDeep reinforcement learning methods → RL for energy, networks and operations
SafeRoPE: Risk-specific Head-wise Embedding Rotation for Safe Generation in Rectified Flow TransformersDiffusion models → Robustness, privacy and adversarial aspects +1
STRIVE: Structured Spatiotemporal Exploration for Reinforcement Learning in Video Question AnsweringMultimodal LLMs / vision-language models → Video understanding with MLLMs +1
DEFT: Distribution-guided Efficient Fine-Tuning for Human AlignmentLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
DriveDreamer-Policy: A Geometry-Grounded World-Action Model for Unified Generation and PlanningWorld models → Embodied and robot world models +1
Hidden Meanings in Plain Sight: RebusBench for Evaluating Cognitive Visual ReasoningLLM reasoning → Multimodal reasoning +1
FourierMoE: Fourier Mixture-of-Experts Adaptation of Large Language ModelsMixture-of-experts → Mixture of LoRA experts
Control-DINO: Feature Space Conditioning for Controllable Image-to-Video DiffusionDiffusion models → Video generation
LiveMathematicianBench: A Live Benchmark for Research-Level Mathematical Reasoning with Proof SketchesLLM reasoning → Mathematical reasoning
From BM25 to Corrective RAG: Benchmarking Retrieval Strategies for Text-and-Table DocumentsRetrieval-augmented generation → RAG pipeline design and evaluation
Causal Scene Narration with Runtime Safety Supervision for Vision-Language-Action DrivingVision-language-action models → VLA for autonomous driving +1
SteerFlow: Steering Rectified Flows for Faithful Inversion-Based Image EditingDiffusion models → Image generation, text-to-image and editing

All 462