# 小米MiMo-V2.6强化学习训练进展公开

- 来源：Fuli Luo (@_LuoFuli)
- 发布时间：2026-09-17 02:52
- AIHOT 分数：43
- AIHOT 链接：https://aihot.news/items/cmu4gvmsw0qearo4wr9i3uujb
- 原文链接：https://x.com/_LuoFuli/status/2100296686719610932

## AI 摘要

小米MiMo-V2.6正处于RL训练中，团队从三方面扩展规模：计算（每步约2B tokens，1568条prompt×16次rollout，全异步）、环境与harness（多任务agentic RL，单次运行混合多种harness）、评分计算（agentic组内信用分配，结合测试用例与rubric奖励）。团队称已研究RL能扩展到多远，将在未来几周逐步开源细节，并直播训练过程。

## 正文

Nearly half a year of silence. We spent it studying one problem: how far RL can scale.

MiMo-V2.6 is in the middle of its RL run right now. Three things we scaled: compute (~2B tokens per step, 1568 prompts × 16 rollouts, fully async), environments and harnesses (multi-task agentic RL, mixed across multiple harnesses in one run), and grader compute (agentic in-group credit assignment, with test-case and rubric-based rewards). We'll open-source the details piece by piece over the coming weeks.

Streaming the run: https://mimo.xiaomi.com/rl/