Papers

New on arXiv

Wed 4 Mar 2026

424 papers

RADIANT-LLM: an Agentic Retrieval Augmented Generation Framework for Reliable Decision Support in Safety-Critical Nuclear EngineeringRetrieval-augmented generation → Agentic RAG, agents and memory +1
Self-Sovereign AgentLLM agents and tool use → General LLM agent frameworks and benchmarks
Prism: Policy Reuse via Interpretable Strategy Mapping in Reinforcement LearningDeep reinforcement learning methods → Interpretable and logic-guided RL
GeoBlock: Inferring Block Granularity from Dependency Geometry in Diffusion Language ModelsDiffusion models → Discrete and text diffusion
Visuospatial Perspective Taking in Multimodal Language ModelsMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
Internal Safety Collapse in Frontier Large Language ModelsLLM safety and security → Safety alignment, refusal and guardrails
Fast and Faithful: Real-Time Verification for Long-Document Retrieval-Augmented Generation SystemsRetrieval-augmented generation → RAG robustness and adaptive retrieval
Beyond Masks: Efficient, Flexible Diffusion Language Models via Deletion-Insertion ProcessesDiffusion models → Discrete and text diffusion
Less is More: Adapting Text Embeddings for Low-Resource Languages with Small Scale Noisy Synthetic DataRetrieval-augmented generation → Generation and context integration +1
Beyond Test-Time Compute Strategies: Advocating Energy-per-Token in LLM InferenceLLM reasoning → Test-time compute and efficient thinking
From Language to Action in Arabic: Reliable Structured Tool Calling via Data-Centric Fine-TuningLLM agents and tool use → Tool learning and function calling
Steering Frozen LLMs: Adaptive Social Alignment via Online Prompt RoutingLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks
Alternating Reinforcement Learning with Contextual Rubric Rewards: Beyond the Scalarization StrategyLLM alignment / RLHF / preference optimisation → Reward hacking and misspecification
DOVA: Deliberation-First Multi-Agent Orchestration for Autonomous Research AutomationLLM agents and tool use → Web, search, and research agents
Auditing Cascading Risks in Multi-Agent Systems via Semantic-Geometric Co-evolutionLLM agents and tool use → Agent threats, guardrails and governance
LightningRL: Breaking the Accuracy-Parallelism Trade-off of Block-wise dLLMs via Reinforcement LearningLLM reasoning → Diffusion reasoning and test-time scaling
Bi-HIL: Bilateral Control-Based Multimodal Hierarchical Imitation Learning via Subtask-Level Progress Rate and Keyframe Memory for Long-Horizon Contact-Rich Robotic ManipulationOffline RL and imitation learning from data → Robot manipulation imitation learning
ChatNeuroSim: An LLM Agent Framework for Automated Compute-in-Memory Accelerator Deployment and OptimizationLLM agents and tool use → Software engineering coding agents
TimeSpot: Benchmarking Geo-Temporal Understanding in Vision-Language Models in Real-World SettingsWorld models → World models for language agents +1
ECHO: Event-Centric Hypergraph Operations via Multi-Agent Collaboration for Multimedia Event ExtractionLLM agents and tool use → Multimodal and embodied agents
Decoding the Pulse of Reasoning VLMs in Multi-Image Understanding TasksLLM reasoning → Multimodal reasoning +1
Neuro-Symbolic Financial Reasoning via Deterministic Fact Ledgers and Adversarial Low-Latency Hallucination DetectorLLM reasoning → Reasoning for agents, planning and code +1
iAgentBench: Benchmarking Sensemaking Capabilities of Information-Seeking Agents on High-Traffic TopicsLLM agents and tool use → Web, search, and research agents
When Sensors Fail: Temporal Sequence Models for Robust PPO under Sensor DriftDeep reinforcement learning methods → Deep RL algorithms and representations
Coordinated Semantic Alignment and Evidence Constraints for Retrieval-Augmented Generation with Large Language ModelsRetrieval-augmented generation → RAG robustness and adaptive retrieval
RoboMME: Benchmarking and Understanding Memory for Robotic Generalist PoliciesVision-language-action models → Reasoning and chain-of-thought in VLA
When Agents Persuade: Rhetoric Generation and Mitigation in LLMsLLM agents and tool use → Social simulation and multi-agent debate +1
SGR3 Model: Scene Graph Retrieval-Reasoning Model in 3DLLM reasoning → Spatial and embodied reasoning
Bootstrapping Exploration with Group-Level Natural Language Feedback in Reinforcement LearningLLM reasoning → Reinforcement learning for reasoning +1
ECG-MoE: Mixture-of-Expert Electrocardiogram Foundation ModelMixture-of-experts → Medical and biomedical MoE
Self-Attribution Bias: When AI Monitors Go Easy on ThemselvesLLM agents and tool use → Software engineering coding agents
Risk-Aware Reinforcement Learning for Mobile ManipulationDeep reinforcement learning methods → RL for robot locomotion and manipulation
Mask-aware inference with State-Space ModelsState-space sequence models (S4, Mamba) → Vision Mamba backbones
Structure-Guided Histopathology Synthesis via Dual-LoRA DiffusionDiffusion models → Medical imaging and segmentation
Latent Particle World Models: Self-supervised Object-centric Stochastic Dynamics ModelingWorld models → Latent predictive world models
Adaptive Memory Admission Control for LLM AgentsLLM agents and tool use → Agent memory systems
Progressive Refinement Regulation for Accelerating Diffusion Language Model DecodingDiffusion models → Discrete and text diffusion
From Spark to Fire: Modeling and Mitigating Error Cascades in LLM-Based Multi-Agent CollaborationLLM agents and tool use → Agent threats, guardrails and governance
Cross-Layer Semantic Flow Reconstruction for Attack Detection in Agentic SystemsLLM agents and tool use → Agent threats, guardrails and governance
A Dual-Helix Governance Approach Towards Reliable Agentic Artificial Intelligence for WebGIS DevelopmentLLM agents and tool use → Multimodal and embodied agents
AgentIR: Reasoning-Aware Retrieval for Deep Research AgentsLLM agents and tool use → Web, search, and research agents
TaxonRL: Reinforcement Learning with Intermediate Rewards for Interpretable Fine-Grained Visual ReasoningLLM reasoning → Multimodal reasoning +1
Robustness of Agentic AI Systems via Adversarially-Aligned Jacobian RegularizationLLM agents and tool use → General LLM agent frameworks and benchmarks
Dual-Modality Multi-Stage Adversarial Safety Training: Robustifying Multimodal Web Agents Against Cross-Modal AttacksLLM agents and tool use → Agent threats, guardrails and governance
ManipulationNet: An Infrastructure for Benchmarking Real-World Robot Manipulation with Physical Skill Challenges and Embodied Multimodal ReasoningLLM reasoning → Spatial and embodied reasoning
Efficient Refusal Ablation in LLM through Optimal TransportLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks +1
A Constrained RL Approach for Cost-Efficient Delivery of Latency-Sensitive ApplicationsDeep reinforcement learning methods → RL for energy, networks and operations
RANGER: Sparsely-Gated Mixture-of-Experts with Adaptive Retrieval Re-ranking for Pathology Report GenerationMixture-of-experts → Medical and biomedical MoE
Underrepresented in Foundation Model Pretraining Data? A One-Shot ProbeMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
Balancing Fidelity, Utility, and Privacy in Synthetic Cardiac MRI Generation: A Comparative StudyDiffusion models → Medical imaging and segmentation
ArtHOI: Articulated Human-Object Interaction Synthesis by 4D Reconstruction from Video PriorsDiffusion models → Video generation
What Does Flow Matching Bring To TD Learning?Diffusion models → Diffusion and flow policies for robotics +1
CAMMSR: Category-Guided Attentive Mixture of Experts for Multimodal Sequential RecommendationMixture-of-experts → MoE for recommender systems
AILS-NTUA at SemEval-2026 Task 12: Graph-Based Retrieval and Reflective Prompting for Abductive Event ReasoningLLM reasoning → Reasoning benchmarks and evaluation
World Properties without World Models: Recovering Spatial and Temporal Structure from Co-occurrence Statistics in Static Word EmbeddingsWorld models → World models for language agents
Dual Diffusion Models for Multi-modal Guided 3D Avatar GenerationDiffusion models → 3D, point cloud and scene generation
V1: Unifying Generation and Self-Verification for Parallel ReasonersLLM reasoning → Test-time compute and efficient thinking
FlowW2N: Whispered-to-Normal Speech Conversion via Flow-MatchingDiffusion models → Audio, speech and music
CubeComposer: Spatio-Temporal Autoregressive 4K 360deg Video Generation from Perspective VideoDiffusion models → 3D, point cloud and scene generation
IPD: Boosting Sequential Policy with Imaginary Planning Distillation in Offline Reinforcement LearningOffline RL and imitation learning from data → Decision transformers and sequence modeling
Memex(RL): Scaling Long-Horizon LLM Agents via Indexed Experience MemoryLLM agents and tool use → Agent memory systems
Agentics 2.0: Logical Transduction Algebra for Agentic Data WorkflowsLLM agents and tool use → Software engineering coding agents
DiverseDiT: Towards Diverse Representation Learning in Diffusion TransformersDiffusion models → Diffusion foundations and sampling
Retrieval or Representation? Reassessing Benchmark Gaps in Multilingual and Visually Rich RAGRetrieval-augmented generation → Multimodal RAG
PRAM-R: A Perception-Reasoning-Action-Memory Framework with LLM-Guided Modality Routing for Adaptive Autonomous DrivingLLM reasoning → Chain-of-thought reasoning for driving
Real5-OmniDocBench: A Full-Scale Physical Reconstruction Benchmark for Robust Document Parsing in the WildMultimodal LLMs / vision-language models → Document understanding and multimodal RAG
Stable and Steerable Sparse Autoencoders with Weight RegularizationMechanistic interpretability → SAE features and their applications
CodeTaste: Can LLMs Generate Human-Level Code Refactorings?LLM agents and tool use → Software engineering coding agents
A Multi-Agent Framework for Interpreting Multivariate Physiological Time SeriesLLM agents and tool use → Medical and scientific domain agents
Unbiased Dynamic Pruning for Efficient Group-Based Policy OptimizationLLM reasoning → Reinforcement learning for reasoning
BeamPERL: Parameter-Efficient RL with Verifiable Rewards Specializes Compact LLMs for Structured Beam Mechanics ReasoningLLM reasoning → Reinforcement learning for reasoning
FastWave: Optimized Diffusion Model for Audio Super-ResolutionDiffusion models → Audio, speech and music
Swimming Under Constraints: A Safe Reinforcement Learning Framework for Quadrupedal Bio-Inspired PropulsionDeep reinforcement learning methods → RL for robot navigation and UAV control
SaFeR: Safety-Critical Scenario Generation for Autonomous Driving Test via Feasibility-Constrained Token ResamplingEnd-to-end autonomous driving → Adversarial scenarios and safety testing
Monitoring Emergent Reward Hacking During Generation via Internal ActivationsLLM alignment / RLHF / preference optimisation → Reward hacking and misspecification
Tuning Just Enough: Lightweight Backdoor Attacks on Multi-Encoder Diffusion ModelsDiffusion models → Robustness, privacy and adversarial aspects +1
Force-Aware Residual DAgger via Trajectory Editing for Precision Insertion with Impedance ControlOffline RL and imitation learning from data → Robot manipulation imitation learning
Self-adapting Robotic Agents through Online Continual Reinforcement Learning with World Model FeedbackWorld models → Latent dynamics models and model-based RL +1
Rethinking the Efficiency and Effectiveness of Reinforcement Learning for Radiology Report GenerationDeep reinforcement learning methods → RL for clinical treatment and surgery
Continuous Modal Logical Neural Networks: Modal Reasoning via Stochastic AccessibilityLLM reasoning → Reasoning benchmarks and evaluation
Training-Free Rate-Distortion-Perception Traversal With DiffusionDiffusion models → Diffusion for compression and wireless
Discriminative Perception via Anchored Description for Reasoning SegmentationLLM reasoning → Multimodal reasoning +1
RIVER: A Real-Time Interaction Benchmark for Video LLMsMultimodal LLMs / vision-language models → Video understanding with MLLMs
Phi-4-reasoning-vision-15B Technical ReportLLM reasoning → Multimodal reasoning
Dual-Solver: A Generalized ODE Solver for Diffusion Models with Dual PredictionDiffusion models → Diffusion foundations and sampling
UniRain: Unified Image Deraining with RAG-based Dataset Distillation and Multi-objective Reweighted OptimizationMixture-of-experts → Vision mixture-of-experts
Structural Action Transformer for 3D Dexterous ManipulationOffline RL and imitation learning from data → Robot manipulation imitation learning
GIPO: Gaussian Importance Sampling Policy OptimizationDeep reinforcement learning methods → RL theory and sample complexity
SCP: Spatial Causal Prediction in VideoLLM reasoning → Spatial and embodied reasoning
Lightweight Visual Reasoning for Socially-Aware RobotsLLM reasoning → Spatial and embodied reasoning
Selecting Offline Reinforcement Learning Algorithms for Stochastic Network ControlOffline RL and imitation learning from data → Offline multi-agent RL
When Safety Becomes a Vulnerability: Exploiting LLM Alignment Homogeneity for Transferable Blocking in RAGRetrieval-augmented generation → Privacy and security of RAG +2
IROSA: Interactive Robot Skill Adaptation using Natural LanguageOffline RL and imitation learning from data → Language-conditioned policies and VLAs
On the Suitability of LLM-Driven Agents for Dark Pattern AuditsLLM agents and tool use → Agent threats, guardrails and governance
Believe Your Model: Distribution-Guided Confidence CalibrationLLM reasoning → Test-time compute and efficient thinking
DeepScan: A Training-Free Framework for Visually Grounded Reasoning in Large Vision-Language ModelsLLM reasoning → Multimodal reasoning +1
Invariance-Based Dynamic Regret MinimizationLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF
SkillVLA: Tackling Combinatorial Diversity in Dual-Arm Manipulation via Skill ReuseVision-language-action models → Human-video and latent-action pretraining
Universal Pansharpening ModelDiffusion models → Inverse problems and image restoration
From Narrow to Panoramic Vision: Attention-Guided Cold-Start Reshapes Multimodal ReasoningLLM reasoning → Multimodal reasoning

All 424