Aligning LLMs by Predicting Preferences from User Writing Samples.

Stphane Aroca-Ouellette, Natalie Mackraz, Barry-John Theobald, Katherine Metcalf

Browse the full ICML paper archive.