Muhao Chen

QLIP: A Dynamic Quadtree Vision Prior Enhances MLLM Performance Without Retraining

FRIEDA: Benchmarking Multi-Step Cartographic Reasoning in Vision-Language Models

Latent Visual Reasoning

Diagnosing and Mitigating Modality Interference in Multimodal Large Language Models

Optimizing Diversity and Quality through Base–Aligned Model Collaboration

Unraveling Indirect In-Context Learning Using Influence Functions

Less Diverse, Less Safe: The Indirect But Pervasive Risk of Test-Time Scaling in Large Language Models

GQA-$\mu$P: The Maximal Parameterization Update for Grouped Query Attention and Fully Sharded Data Parallel

Towards Policy-Compliant Agents: Learning Efficient Guardrails for Policy Violation Detection

ArenaBencher: Automatic Benchmark Evolution via Multi-Model Competitive Evaluation

Unbiased Visual Reasoning with Controlled Visual Inputs

Is Extending Modality The Right Path Towards Omni-Modality?

LayerIF: Estimating Layer Quality for Large Language Models using Influence Functions

BadJudge: Backdoor Vulnerabilities of LLM-As-A-Judge

Benchmarking Vision Language Model Unlearning via Fictitious Facial Identity Dataset

MuirBench: A Comprehensive Benchmark for Robust Multi-image Understanding

Familiarity-Aware Evidence Compression for Retrieval-Augmented Generation

Unraveling Cross-Modality Knowledge Conflicts in Large Vision-Language Models