Fetching the paper…

Direct Large Language Model Alignment Through Self-Rewarding Contrastive Prompt Distillation · Around