AI.info

Yankai Lin

Explore Yankai Lin on AI.info.

• Less Noise, More Voice: Reinforcement Learning for Reasoning via Instruction Purification
• DARC: Decoupled Asymmetric Reasoning Curriculum for LLM Evolution
• Forest Before Trees: Latent Superposition for Efficient Visual Reasoning
• LaSeR: Reinforcement Learning with Last-Token Self-Rewarding