Self-Training with Direct Preference Optimization Improves Chain-of-Thought Reasoning Paper • 2407.18248 • Published Jul 25 • 31
