COLING 2025 - Cross-Refine:Improving Natural Language Explanation Generation by Learning in Tandem

Qianli Wang, a member of the XplaiNLP research group at Technische Universität Berlin and Deutsches Forschungszentrum für Künstliche Intelligenz (DFKI), has been accepted at COLING 2025! 🎉

📝 Cross-Refine: Improving Natural Language Explanation Generation by Learning in Tandem (https://lnkd.in/dbFHqiNk)

In this paper, we introduce an novel approach Cross-Refine where two LLMs work together to generate/refine the natural language explanations.

A big thanks to all my co-authors and advisors Tatiana Anikina, Nils Feldhus, Simon Ostermann, Sebastian Möller, and Dr. Vera Schmitt.