Wenxuan Zhang

RAVENEA: A Benchmark for Multimodal Retrieval-Augmented Visual Culture Understanding

PEAR: Phase Entropy Aware Reward for Efficient Reasoning

SpaCE-Eval: A Benchmark for Real-World Multi-Modal Reasoning

Difficulty–Diversity Collaborative Filtering for Data-Efficient LLM Fine-Tuning

When Less Language is More: Language-Reasoning Disentanglement Makes LLMs Better Multilingual Reasoners

Assessing Judging Bias in Large Reasoning Models: An Empirical Study

The Rise of Parameter Specialization for Knowledge Storage in Large Language Models

The Emergence of Abstract Thought in Large Language Models Beyond Any Language

Understanding and Enhancing Safety Mechanisms of LLMs via Safety-Specific Neuron

Auto-Arena: Automating LLM Evaluations with Agent Peer Battles and Committee Discussions

From General to Expert: Custom Pruning LLMs Across Language, Domain, and Task