Papers

New on arXiv

Sat 27 Jun 2026

206 papers

Beyond Accuracy: A Qualitative Analysis of Vision-Language Models for Hate Speech Detection in MemesMultimodal LLMs / vision-language models → Social bias and fairness in VLMs
Truth Lies Deep: Countering Semantic Camouflage via Latent Intent VerificationLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks +1
NEXT: Reasoning-Driven Video Recommendation via a Vision-Language ModelLLM reasoning → Multimodal reasoning +1
Between Suppression and Collapse: Evaluating Narrative Unlearning with LENSLLM safety and security → Machine unlearning
Learning Spatio-Temporal Foundation Models from Pure Synthetic DataSynthetic data and self-improvement for LLMs → Model collapse from synthetic data
An Agentic Interface for End-to-End Probabilistic Seismic Hazard and Risk AnalysisLLM agents and tool use → Multimodal and embodied agents
High-accuracy Low-Bit KV-Cache Quantization via Local Distribution RestorationEfficient LLM inference and serving → KV cache quantization
Characterizing Large Language Model Agentic Workflows: A Study on N8n EcosystemLLM agents and tool use → Software engineering coding agents
A Novel Latent-Class Attack and its Detection by Class Subspace OrthogonalizationLLM safety and security → Backdoor and data poisoning attacks
TrafficAlign: Aligning Large Language Models for Traffic Scenario GenerationLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF
HorizonRelight: Relighting Long-horizon Videos Consistently via Diffusion TransformersDiffusion models → 3D, point cloud and scene generation
DiLaServe: High SLO Attainment Serving for Diffusion Language ModelsDiffusion models → Discrete and text diffusion
AB-RAG: Adaptive Budgeted Retrieval-Augmented Generation for Reliable Question AnsweringRetrieval-augmented generation → RAG robustness and adaptive retrieval
TAP-VLA: Tactile Annotation Prompting for Vision Language Action ModelsVision-language-action models → Tactile, force-aware and humanoid VLA
From Tool Connection to Execution Control: Benchmarking Security Invariants in MCP-Style Agent RuntimesLLM agents and tool use → Agent threats, guardrails and governance
ThinkProbe: Beyond Accuracy -- Structural Profiling of Open-Ended LLM Reasoning Traces via Non-Generative Thought GraphsLLM reasoning → Test-time compute and efficient thinking
x-Prediction Flow: Efficient Continuous Decoding for Masked Diffusion Language ModelsDiffusion models → Discrete and text diffusion
Flow Matching in Feature Space for Stochastic World ModelingWorld models → Video-generation world models and simulators +1
Human-in-the-Loop Nugget Annotation for Accountable LLM-as-a-Judge EvaluationsLLM agents and tool use → General LLM agent frameworks and benchmarks
Memory as an Attack Surface in LLM Agents: A Study on Multiple-Choice Question AnsweringLLM agents and tool use → Agent memory systems
Preventing Error Propagation in Multi-Agent AI through Runtime MonitoringLLM reasoning → Reasoning for agents, planning and code +1
Customized Generative AI Agent for Transportation Engineering Practice: A Development and Continued Pre-training GuidelineLLM agents and tool use → General LLM agent frameworks and benchmarks
Reward-Free Code Alignment from Pretrained or Fine-Tuned LLM: Unpacking the Trade-offs for Code GenerationLLM alignment / RLHF / preference optimisation → Process reward models for reasoning
Bad company corrupts good morals: Understanding and Measuring Narrative-Induced Moral Reasoning Degradation in LLMsLLM safety and security → Safety alignment, refusal and guardrails
Self-Evolving Agentic Image Restoration via Deliberate Planning and Intuitive ExecutionLLM agents and tool use → Multimodal and embodied agents +1
FlipGuard: Defending Large Language Models Against Quantization-Conditioned Backdoor AttacksLLM safety and security → Backdoor and data poisoning attacks
When Latent Agents Lie: KV-Cache Integrity in Multi-Agent LLM CollaborationLLM agents and tool use → Software engineering coding agents
Modification-Considering Value Learning for Reward Hacking Mitigation in RLLLM alignment / RLHF / preference optimisation → Reward hacking and misspecification +1
A3M: Adaptive, Adversarial and Multi-Objective Learning for Strategic Bidding in Repeated AuctionsDeep reinforcement learning methods → Adversarial robustness and RL in finance
ReGuide: From Test-Time Guidance to Self-Improving Diffusion PoliciesDiffusion models → Diffusion and flow policies for robotics
EVLA: An Electro-Aware Multimodal Assistant for Physically-Grounded Driving Reasoning and ControlLLM reasoning → Chain-of-thought reasoning for driving +2
Multi-Agent Routing as Set-Valued Prediction: A WildChat Benchmark and Cost-Aware EvaluationLLM agents and tool use → General LLM agent frameworks and benchmarks
Cross-Sensor SAR Data Generation Using Diffusion Models and Feature MigrationDiffusion models → Inverse problems and image restoration
PASTA: A Paraphrasing And Self-Training Approach for Knowledge Updating in LLMsSynthetic data and self-improvement for LLMs → Synthetic reasoning data and bootstrapping
A Task-Driven and Quality-Assured Agent Framework for SAR Data GenerationLLM agents and tool use → Multimodal and embodied agents
On Test-Time Scaling for Vision-Language ModelsLLM reasoning → Multimodal reasoning +1
Personalizing MLLMs via Reinforced Multimodal Reference GameMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
LAMP: Lean-based Agentic framework with MCP and Proof RepairLLM reasoning → Mathematical reasoning +1
The Contagion Tensor: A Framework for Measuring Output-Distribution Coupling in Multi-Agent LLM Systems -- and Auditing the Claims It EnablesLLM agents and tool use → General LLM agent frameworks and benchmarks
Fisher-Routed Mixture of Experts for Federated Class-Incremental LearningMixture-of-experts → Federated mixture-of-experts
Q-DASC: State-of-the-Art Safe Quantum Control for HVAC under Local Model MisspecificationDeep reinforcement learning methods → Quantum reinforcement learning
HARD-KV: Head-Adaptive Regularization for Decoding-time KV CompressionEfficient LLM inference and serving → KV cache optimization
Labeling Training Data for Entity Matching Using Large Language ModelsKnowledge distillation → Knowledge distillation for LLMs
ViPSim: Collaborating Visual and Parameter Spaces for Consistent Long-Horizon Embodied World ModelsWorld models → Embodied and robot world models +1
From Determinism to Delegation: AI-Native Software Engineering and the Evolution of the Agentic EngineerLLM agents and tool use → Software engineering coding agents
Stochastic Optimal Control Sampling for Diffusion Inverse ProblemsDiffusion models → Inverse problems and image restoration
Multimodal Graph RAG for Long-range Visually Rich Document UnderstandingMultimodal LLMs / vision-language models → Document understanding and multimodal RAG +1
Mechanistic Personality Analysis of LLMs: Steering Personality via Latent Feature InterventionsMechanistic interpretability → Activation steering methods
Vision-Language Models for Deployable Social Robot Navigation: Bridging Semantic Reasoning and Low-Level ControlLLM reasoning → Spatial and embodied reasoning +1
X-Mind: Efficient Visual Chain-of-Thought via Predictive World Model for End-to-End DrivingWorld models → World models for autonomous driving +3
A Physics-Grounded Benchmark for Multi-Agent Dynamics in World ModelsWorld models → Latent predictive world models
A Path-Space Formulation of Prediction in World Models: From a Single Action to Prediction, Planning, and IrreversibilityWorld models → Latent predictive world models
Self-Supervised Theorem Discovery in a Formal Axiomatic SystemLLM reasoning → Mathematical reasoning
FreqOrtho-SR: Frequency-Guided Orthogonal Expert Learning for Real-World Image Super-ResolutionDiffusion models → Inverse problems and image restoration
Agent Safety Is Action AlignmentLLM agents and tool use → Agent threats, guardrails and governance
5ting at SemEval-2026 Task 8: Strong End-to-End Multi-Turn RAG via LLM-Based Reranking and Faithfulness ControlRetrieval-augmented generation → RAG pipeline design and evaluation
Agentic Abstention: Do Agents Know When to Stop Instead of Act?LLM agents and tool use → General LLM agent frameworks and benchmarks
CCRC: A Change-Aware Captioning and Reasoning Chain for Image Change Captioning and SegmentationLLM reasoning → Multimodal reasoning
ComMem: Complementary Memory Systems for Test-Time Adaptation of Vision-Language ModelsMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
J-LAW: Joint Localization and Action-Conditioned World Modeling via Coupled Latent Factor GraphsWorld models → Latent predictive world models
The Two Genie Game: Adoption and Welfare in Audit-Grounded AI GovernanceLLM alignment / RLHF / preference optimisation → Human values and AI feedback alignment
BV-Blend: Uncertainty-Weighted Historical Baselines for Stable Critic-Free RL with Verifiable RewardsLLM reasoning → Reinforcement learning for reasoning +1
An AI agent for treatment reasoning over a biomedical tool universeLLM reasoning → Clinical and biomedical applications +1
Capability Gates Are Not Authorization: Confused-Deputy Failures in LLM Agent FrameworksLLM agents and tool use → Agent threats, guardrails and governance
SATB-VR: Training Few-Step Video Restoration Diffusion Model using SNR-Aware Trajectory BlendingDiffusion models → Inverse problems and image restoration
Constrained Tabular Diffusion for FinanceDiffusion models → Flow matching, time series and science
Entropy-Regularized Reinforcement Learning for Linear-Quadratic Stackelberg Differential Games in Regime-Switching Diffusion ModelsDeep reinforcement learning methods → RL theory and sample complexity
Entropy Regularized Reinforcement Learning for Zero-Sum Stochastic Differential Games in a Regime-Switching Jump-Diffusion ProcessDeep reinforcement learning methods → RL theory and sample complexity
Why Trust Your Agent? Empirical Security Gains from TRiSM-Guided Agentic Workflows in HealthcareLLM agents and tool use → Agent threats, guardrails and governance
When More Sampling Hurts: The Modal Ceiling and Correlation Ceiling of Test-Time ScalingLLM reasoning → Test-time compute and efficient thinking
Deep Learning for Reflected BSDEs: Regularization and Error Analysis
CDPR: Counterfactual Advantage-based Credit Assignment for Cost-Aware Sequential Medical Diagnosis
Forecasting Weather-Driven Price Dynamics Across Sri Lankan Tea Market Catalogues
Data Predictability Shapes Weibull Weight-Scale Growth in Transformer Training
Modular Cognitive Architecture Emerges in Large Language Models
JaleesBench: Are AI Assistants Good Spiritual Company?
Need or Reputation? Region, National Income, and Academic Status in LLM Gatekeeping Decisions
Monsoon Mayhem to Market Waves: Forecasting Fisheries Resilience in Sri Lanka
StanceBench: A Benchmark for Audio LLM-Based Interpersonal Stance Evaluation from Speech
Feature Generation Using LLMs: An Evolutionary Algorithm Approach
More Than Memory: Task-Conditioned Signed FFN Writes in Long-Context Retrieval
Comprehensive Evaluation of Machine Learning for Type 2 Diabetes Risk Prediction: Large-Scale External Validation and Fairness Analysis
SOS-LoRA: Static Orthogonal-Subspace Low-Rank Adaptation with Fixed Multi-Scale Scaling
Benchmarking Machine Learning Models for Multi-Omics-Based Breast Cancer Prediction
Uncertainty-Aware Sequential Decision Rules for Event-Triggered LLM Invocation in Streaming Systems
The Ramanujan Challenge For AI
Do Diabetic Foot Ulcer Segmentation Models Generalize? A Cross-Dataset Benchmark of CNN and Transformer Architectures
Reliability-Aware Monocular Depth Supervision for Sparse-View Neural Reconstruction
ALM2Vec: Learning Audio Embeddings for Universal Audio Retrieval with Large Audio-Language Models
When Stopping Fails: Rethinking Minimal Risk Conditions through Human-Interactive Autonomous Driving for Safe Transportation Systems
Multivariate Varying-Coefficient BART with Graphical Horseshoe Priors
LLM Semantic Signaling Game and Mechanism Design: Systematic Blindness, Awareness Shaping, and Mindset Dynamics
Managing the Human Fallback: Skill Investment Under Improving AI and Worker Mobility
Few-Step Boltzmann Generators via Scalable Likelihood Flow Maps
A Deep Multiscale Neural Network for Accurate Neurological Disorder Detection from MRI Scans and Real-Time Web Deployment
spca: An R package to Compute Least Squares Sparse Principal Components
BTI-Net: Bidirectional Decoder-Level Task Interaction via Uncertainty-Aware Gating for Multi-Task Medical Image Analysis
Unified Complex-valued Neural Network: A Magnitude-Phase Computational Model for Event-Driven Neuromorphic Learning
Connectivity Estimation using Stochastic Graph Heat Modelling
From Fog Chamber to Aircraft Window: Pixel-Registered Imaging and Synthetic Fine-Tuning Enable Cross-Domain Defogging

All 206