Alekh Agarwal

A new method enhances language models by focusing on user preferences.

Introducing a framework for fine-tuning language models to balance multiple objectives.

Teaching machines to learn without revealing expert secrets is crucial for privacy.