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

Open in new window