BeaverTails: Towards Improved Safety Alignment of LLM via a Human-Preference Dataset Paper • 2307.04657 • Published Jul 10, 2023 • 6
