Hao AI Lab· @haoailab · X·· 9 天前AI 评分62
AI 导读
Hao AI Lab 提出了一种多智能体分工策略,即让快速的智能体负责跑腿,让强大的推理者负责思考。作者同时分享了 200 个测试案例的回放链接以及项目代码。
正文 · 原文
5️⃣ The moral: let fast agents do the legwork and a strong reasoner do the thinking. One genius, many minds. 🧠👣 Replay all 200 cases: https://snyhlx.github.io/the_irregulars_detective_game/results.html Code, as well as more gaming case study for agent swarm behaviors: https://github.com/lmgame-org/Gaming-JevSwarm Happy sleuthing!
来源:Hao AI Lab · x.com