Papers

New on arXiv

Tue 31 Mar 2026

447 papers

KPI2KVI: A Multi Agent Workflow for Calculating Key Value Indicators from Service DescriptionsLLM agents and tool use → General LLM agent frameworks and benchmarks
Train the Trainers -- An Agentic AI Framework for Peer-Based Mental Health Support in Battlefield EnvironmentsLLM agents and tool use → Medical and scientific domain agents
Helping Customers in Distress: An LLM-powered Agent that Converses, Probes, and RoutesLLM agents and tool use → General LLM agent frameworks and benchmarks
Beyond Individual Mimicry: Constructing Human-Like Social network with Graph-Augmented LLM AgentsLLM agents and tool use → Social simulation and multi-agent debate
LLM Agents Predict Social Media Reactions but Do Not Outperform Text Classifiers: Benchmarking Simulation Accuracy Using 120K+ Personas of 1511 HumansLLM agents and tool use → Social simulation and multi-agent debate
Can LLMs Infer Conversational Agent Users' Personality Traits from Chat History?LLM agents and tool use → Social simulation and multi-agent debate
ICAT: Incident-Case-Grounded Adaptive Testing for Physical-Risk Prediction in Embodied World ModelsWorld models → World models for navigation
GraphRAG-Router: Learning Cost-Efficient Routing over GraphRAGs and LLMs with Reinforcement LearningLLM reasoning → Knowledge- and tool-augmented reasoning +1
CoLLM: Continuous Adaptation for SLO-Aware LLM Serving on Shared GPU ClustersEfficient LLM inference and serving → Serving systems and request scheduling
AlphaLab: Autonomous Multi-Agent Research Across Optimization Domains with Frontier LLMsLLM agents and tool use → General LLM agent frameworks and benchmarks
V-Reflection: Transforming MLLMs from Passive Observers to Active InterrogatorsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
KAIJU: An Executive Kernel for Intent-Gated Execution of LLM AgentsLLM agents and tool use → General LLM agent frameworks and benchmarks +1
Collaborative AI Agents and Critics for Fault Detection and Cause Analysis in Network TelemetryLLM agents and tool use → Software engineering coding agents
Prompt-Guided Prefiltering for VLM Image CompressionMultimodal LLMs / vision-language models → Visual token pruning and efficiency
CASA: Classification Augmented with Safety Attention for Robust Multimodal AlignmentLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks +1
Asymmetric Actor-Critic for Multi-turn LLM AgentsLLM reasoning → Reasoning for agents, planning and code +1
When is Generated Code Difficult to Comprehend? Assessing AI Agent Python Code Proficiency in the WildLLM agents and tool use → Software engineering coding agents
SANA I2I: A Text Free Flow Matching Framework for Paired Image to Image Translation with a Case Study in Fetal MRI Artifact ReductionDiffusion models → Medical imaging and segmentation
SYNTHONY: A Stress-Aware, Intent-Conditioned Agent for Deep Tabular Generative Models SelectionDiffusion models → Flow matching, time series and science +1
MambaVoiceCloning: Efficient and Expressive Text-to-Speech via State-Space Modeling and Diffusion ControlState-space sequence models (S4, Mamba) → Speech and audio SSMs
Improvisational Games as a Benchmark for Social Intelligence of AI Agents: The Case of ConnectionsLLM reasoning → Reasoning for agents, planning and code +1
Data-Driven Reachability Analysis via Diffusion Models with PAC GuaranteesDiffusion models → Diffusion and flow policies for robotics
The Geometry of Compromise: Unlocking Generative Capabilities via Controllable Modality AlignmentMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
Structure- and Event-Driven Frameworks for State Machine Modeling with Large Language ModelsLLM reasoning → Reasoning for agents, planning and code
OmniSch: A Multimodal PCB Schematic Benchmark For Structured Diagram Visual ReasoningLLM reasoning → Multimodal reasoning
Autonomous Adaptive Solver Selection for Chemistry Integration via Reinforcement LearningDeep reinforcement learning methods → RL for energy, networks and operations
Can Large Language Models Self-Correct in Medical Question Answering? An Exploratory StudyLLM reasoning → Clinical and biomedical applications
A Safety-Aware Role-Orchestrated Multi-Agent LLM Framework for Behavioral Health Communication SimulationLLM agents and tool use → Medical and scientific domain agents
REM-CTX: Automated Peer Review via Reinforcement Learning with Auxiliary ContextDeep reinforcement learning methods → RL training of language and agent models
Hierarchical Discrete Flow Matching for Graph GenerationDiffusion models → Flow matching, time series and science
MAC-Attention: a Match-Amend-Complete Scheme for Fast and Accurate Attention ComputationEfficient LLM inference and serving → KV cache optimization
Do Language Models Know When They'll Refuse? Probing Introspective Awareness of Safety BoundariesLLM safety and security → Safety alignment, refusal and guardrails
Diversity-Aware Reverse Kullback-Leibler Divergence for Large Language Model DistillationKnowledge distillation → Knowledge distillation for LLMs
Similarity of Neural Network Representations in SuperpositionMechanistic interpretability → Sparse autoencoder training and evaluation
DreamControl-v2: Simpler and Scalable Autonomous Humanoid Skills via Trainable Guided Diffusion PriorsDiffusion models → Diffusion and flow policies for robotics
Offline Constrained RLHF with Multiple Preference OraclesLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF
Making Sense of AI Agents Hype: Adoption, Architectures, and Takeaways from PractitionersLLM agents and tool use → Software engineering coding agents
Explainable AI for Blind and Low-Vision Users: Navigating Trust, Modality, and Interpretability in the Agentic EraLLM agents and tool use → Multimodal and embodied agents
Agentic AI and Occupational Displacement: A Multi-Regional Task Exposure Analysis of Emerging Labor Market DisruptionLLM agents and tool use → General LLM agent frameworks and benchmarks
Q-Mask: Query-driven Causal Masks for Text Anchoring in OCR-Oriented Vision-Language ModelsMultimodal LLMs / vision-language models → Document understanding and multimodal RAG
Open, Reliable, and Collective: A Community-Driven Framework for Tool-Using AI AgentsLLM agents and tool use → Tool learning and function calling
Oblivion: Self-Adaptive Agentic Memory Control through Decay-Driven ActivationLLM agents and tool use → Agent memory systems
Hierarchical Chain-of-Thought: Enhancing LLM Reasoning Performance and EfficiencyLLM reasoning → Prompting and chain-of-thought reasoning
RawGen: Learning Camera Raw Image GenerationDiffusion models → Inverse problems and image restoration
One Panel Does Not Fit All: Case-Adaptive Multi-Agent Deliberation for Clinical PredictionLLM agents and tool use → Medical and scientific domain agents
PASM: Population Adaptive Symbolic Mixture-of-Experts Model for Cross-location Hurricane Evacuation Decision PredictionMixture-of-experts → MoE for time series and graphs
Terminal Agents Suffice for Enterprise AutomationLLM agents and tool use → Software engineering coding agents
Evolution Strategies for Deep RL pretrainingDeep reinforcement learning methods → RL for robot locomotion and manipulation
Towards Automatic Soccer Commentary Generation with Knowledge-Enhanced Visual ReasoningLLM reasoning → Multimodal reasoning
Generalizable Dense Reward for Long-Horizon Robotic TasksOffline RL and imitation learning from data → Language-conditioned policies and VLAs
The Energy Footprint of LLM-Based Environmental Analysis: LLMs and Domain ProductsRetrieval-augmented generation → Agentic RAG, agents and memory
Video Models Reason Early: Exploiting Plan Commitment for Maze SolvingLLM reasoning → Multimodal reasoning +1
Aligned, Orthogonal or In-conflict: When can we safely optimize Chain-of-Thought?LLM reasoning → Prompting and chain-of-thought reasoning
EnsembleSHAP: Faithful and Certifiably Robust Attribution for Random Subspace MethodLLM safety and security → Membership inference attacks
Cognitive Friction: A Decision-Theoretic Framework for Bounded Deliberation in Tool-Using AgentsLLM agents and tool use → Tool learning and function calling
Architecting Secure AI Agents: Perspectives on System-Level Defenses Against Indirect Prompt Injection AttacksLLM agents and tool use → Agent threats, guardrails and governance +1
Tracking Equivalent Mechanistic Interpretations Across Neural NetworksMechanistic interpretability → Circuits and mechanistic analysis
Enhancing Structural Mapping with LLM-derived Abstractions for Analogical Reasoning in NarrativesLLM reasoning → Reasoning benchmarks and evaluation
Extending MONA in Camera Dropbox: Reproduction, Learned Approval, and Design Implications for Reward-Hacking MitigationLLM alignment / RLHF / preference optimisation → Reward hacking and misspecification
Think Anywhere in Code GenerationLLM reasoning → Reasoning for agents, planning and code
Diagnosing Long-Video Quantitative Reasoning in Multimodal LLMs via Enumeration and CountingLLM reasoning → Multimodal reasoning +1
SkillReducer: Optimizing LLM Agent Skills for Token EfficiencyLLM agents and tool use → Agent skill learning and evolution
C-TRAIL: A Commonsense World Framework for Trajectory Planning in Autonomous DrivingLLM reasoning → Chain-of-thought reasoning for driving +1
ATP-Bench: Towards Agentic Tool Planning for MLLM Interleaved GenerationMultimodal LLMs / vision-language models → Document understanding and multimodal RAG +1
ShapE-GRPO: Shapley-Enhanced Reward Allocation for Multi-Candidate LLM TrainingLLM reasoning → Reinforcement learning for reasoning
An Output Feedback Q-learning Algorithm for Optimal Control of Nonlinear Systems with Koopman Linear EmbeddingDeep reinforcement learning methods → RL theory and sample complexity
DIAL: Decoupling Intent and Action via Latent World Modeling for End-to-End VLAWorld models → Embodied and robot world models +2
SceneTeract: Probing and Improving Agent-Aware Activity Reasoning in 3D Indoor ScenesLLM agents and tool use → Multimodal and embodied agents
Training-Free Dynamic Upcycling of Expert Language ModelsMixture-of-experts → MoE large language models
TSHA: A Benchmark for Visual Language Models in Trustworthy Safety Hazard Assessment ScenariosMultimodal LLMs / vision-language models → Chart understanding and anomaly detection
SHIFT: Stochastic Hidden-Trajectory Deflection for Removing Diffusion-based WatermarkLLM safety and security → Watermarking generated content and models
BotVerse: Real-Time Event-Driven Simulation of Social AgentsLLM agents and tool use → Social simulation and multi-agent debate
Editing on the Generative Manifold: A Theoretical and Empirical Study of General Diffusion-Based Image Editing Trade-offsDiffusion models → Image generation, text-to-image and editing
Unveiling the Reasoning Process of Large Language ModelsLLM reasoning → Reasoning benchmarks and evaluation
Symphony for Medical Coding: A Next-Generation Agentic System for Scalable and Explainable Medical CodingLLM agents and tool use → Medical and scientific domain agents
Client-Verifiable and Efficient Federated Unlearning in Low-Altitude Wireless NetworksLLM safety and security → Federated unlearning and poisoning
A Comprehensive Information-Decomposition Analysis of Large Vision-Language ModelsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Near-Miss: Latent Policy Failure Detection in Agentic WorkflowsLLM agents and tool use → Software engineering coding agents
CutClaw: Agentic Hours-Long Video Editing via Music SynchronizationLLM agents and tool use → Multimodal and embodied agents
6GAgentGym: Tool Use, Data Synthesis, and Agentic Learning for Network ManagementLLM agents and tool use → Agentic AI for wireless networks +1
An Empirical Study of Multi-Agent Collaboration for Automated ResearchLLM agents and tool use → Web, search, and research agents
Unify-Agent: A Unified Multimodal Agent for World-Grounded Image SynthesisLLM agents and tool use → Multimodal and embodied agents
Video-Oasis: Rethinking Evaluation of Video UnderstandingLLM reasoning → Multimodal reasoning
Learning Diagnostic Reasoning for Decision Support in ToxicologyLLM reasoning → Clinical and biomedical applications
FlowID : Enhancing Forensic Identification with Latent Flow-Matching ModelsDiffusion models → Image generation, text-to-image and editing
Emotion Diffusion Classifier with Adaptive Margin Discrepancy Training for Facial Expression RecognitionDiffusion models → Audio, speech and music
Total Variation Guarantees for Sampling with Stochastic LocalizationDiffusion models → Diffusion foundations and sampling
Stand-Alone Complex or Vibercrime? Exploring the adoption and innovation of GenAI tools, coding assistants, and agents within cybercrime ecosystemsLLM agents and tool use → Agent threats, guardrails and governance
Can LLM Agents Identify Spoken Dialects like a Linguist?LLM agents and tool use → General LLM agent frameworks and benchmarks
GraSP-STL: A Graph-Based Framework for Zero-Shot Signal Temporal Logic Planning via Offline Goal-Conditioned Reinforcement LearningOffline RL and imitation learning from data → Robot learning from demonstration
TrafficMoE: Heterogeneity-aware Mixture of Experts for Encrypted Traffic ClassificationMixture-of-experts → Expert routing strategies
Target-Aligned Reinforcement LearningDeep reinforcement learning methods → Deep RL algorithms and representations
Learning to Generate Formally Verifiable Step-by-Step Logic Reasoning via Structured Formal IntermediariesLLM reasoning → Reinforcement learning for reasoning
MemFactory: Unified Inference & Training Framework for Agent MemoryLLM agents and tool use → Agent memory systems
Calibrated Confidence Expression for Radiology Report GenerationMultimodal LLMs / vision-language models → Medical vision-language models
iPoster: Content-Aware Layout Generation for Interactive Poster Design via Graph-Enhanced Diffusion ModelsDiffusion models → Image generation, text-to-image and editing
M-MiniGPT4: Multilingual VLLM Alignment via Translated DataLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
SeGPruner: Semantic-Geometric Visual Token Pruner for 3D Question AnsweringMultimodal LLMs / vision-language models → Embodied and 3D vision-language models +1
Seeing the Evidence, Missing the Answer: Tool-Guided Vision-Language Models on Visual IllusionsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Diffusion-Guided Cooperative Policy Learning for Target Tracking Based on Underwater Mobile Agent NetworksDeep reinforcement learning methods → RL for robot navigation and UAV control

All 447