Wei Ye

4 papers Tongji University

• Boosting Resilience of Large Language Models through Causality-Driven Robust Optimization
• Preference-driven Knowledge Distillation for Few-shot Node Classification
• SAEMark: Steering Personalized Multilingual LLM Watermarks with Sparse Autoencoders
• VLM-R³: Region Recognition, Reasoning, and Refinement for Enhanced Multimodal Chain-of-Thought