Breaking Mental Set to Improve Reasoning through Diverse Multi-Agent Debate

Yexiang Liu, Jie Cao, Zekun Li, Ran He, Tieniu Tan
2/12/2026

No official code found. Best implementation: MraDonkey/DMAD (20 stars, Apache-2.0, last commit 2025-04-22).

Semantic Scholar

Code Implementations(1)

MraDonkey/DMADCommunity72%
Apache-2.0

[ICLR 2025] Breaking Mental Set to Improve Reasoning through Diverse Multi-Agent Debate

202PythonMar 27, 20251 years agoApache-2.0
debatellmsmllmsmulti-agentmultiagent+3 more

Cite this paper

@article{liu2026breaking,
  title  = {Breaking Mental Set to Improve Reasoning through Diverse Multi-Agent Debate},
  author = {Yexiang Liu and Jie Cao and Zekun Li and Ran He and Tieniu Tan},
  year   = {2026},
  url    = {https://api.semanticscholar.org/CorpusID:fb9759e2e360c64015f4c4794505a79a768d0282},
  journal = {ICLR 2025 2025}
}

Discussion