AI.info

Zhengxu Yu

Explore Zhengxu Yu on AI.info.

• Memento 3: Model-Based Recursive Self-Improvement through Reflective Rulebooks
• TokenSqueeze: Performance-Preserving Compression for Reasoning LLMs