(ACL 2024) Rethinking the Bounds of LLM Reasoning: Are Multi-Agent Discussions the Key? - View it on GitHub