Muning Wen

A look at how tree-search methods improve large language models.

TRAD improves how agents select and use examples for better decisions.

AGDC combines reinforcement learning and self-feedback for better gas leak estimation.