Hugh Zhang

A new method allows language models to learn addition skills independently.

A new algorithm merges BQL and CFR for better learning in dynamic environments.

Research uncovers concerns about large language models' math reasoning skills.

A novel approach to improve language models using human feedback.