AI.info

Tianyi Lin

Explore Tianyi Lin on AI.info.

• A Zeroth-Order Paradigm for LLM Preference Alignment
• R1-SyntheticVL: Is Synthetic Data from Generative Models Ready for Multimodal Large Language Model?
• Reward-free Alignment for Conflicting Objectives
• Exploration vs Exploitation: Rethinking RLVR through Clipping, Entropy, and Spurious Reward