Rishabh Joshi

Research aims to make language models safer and more useful for users.

A fresh approach to training reward models enhances AI alignment with human preferences.

This method helps AIs learn through creating and solving challenges.