MraDonkey/DMAD
[ICLR 2025] Breaking Mental Set to Improve Reasoning through Diverse Multi-Agent Debate
Implements multi-agent debate where agents employ distinct reasoning strategies (Chain-of-Thought, Step-Back Prompting, Compositional CoT, etc.) rather than homogeneous approaches, allowing agents to learn from alternative problem-solving perspectives and collectively correct errors. Evaluated on math, chemistry, physics, and multimodal benchmarks using both standard LLMs and vision-language models like LLaVA, showing consistent improvements over single-strategy debate in fewer debate rounds.
No commits in the last 6 months.
Stars
19
Forks
2
Language
Python
License
Apache-2.0
Category
Last pushed
Apr 22, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/agents/MraDonkey/DMAD"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
Detrol/quorum-cli
Multi-agent AI discussion CLI for structured debates between LLMs
starshine-f/Agent-Debate
A multi-agent debate framework supporting AI-vs-AI and Human-vs-AI modes with customizable...
BondarenkoCom/hexnest-mvp-showcase
AI debate arena — agents argue, run Python mid-debate, discover via MCP. Live:...
kevincollinsirl/startup-strategy-council
Multi-agent AI strategic decision system for startups. Get recommendations from 6 specialized AI...
XueJourney/AgentRound
🗣️ AgentRound — 多模型 AI 圆桌讨论框架。让多个 AI 模型围绕同一主题展开多轮辩论，支持并发调用、人类实时介入、智能 Token 裁剪、Rich 终端渲染和...