研究者提出Octopus多智能体强化学习框架

1. 13:00
   arXiv 机器学习

   Octopus：面向隐式 KV 多智能体强化学习的“单脑多分支”框架

   研究者提出单模型多智能体强化学习框架 Octopus，多线程共享同一骨干网络并通过 GPU 驻留的 KV 状态 DAG 进行协同。框架结合零拷贝 Radix Forking、Attention-Only KV Mounting 与 Packed DAG-GRPO，将离题干扰降低 21 倍。

Current heat 8·Comparable peak 10（Nov 6 14:00）·Comparable change, last 24 hours–

Trends compare only the same parties observed continuously, so the scope can be narrower than current heat counts. Move the pointer or tap the chart to see hourly heat; use the left and right arrow keys on a keyboard.