Cross-Refine: Improving Natural Language Explanation Generation by Learning in Tandem

Open in new window