Self-Training with Direct Preference OptimizationImproves Chain-of-Thought Reasoning 논문리뷰

https://arxiv.org/pdf/2407.18248