Can ChatGPT Defend its Belief in Truth? Evaluating LLM Reasoning via Debate.

Boshi Wang, Xiang Yue, Huan Sun

Browse the full EMNLP paper archive.