Xiaomi's CodeMidas builds 5,545 coding RL environments from raw source code
XiaomiMiMo · hf · 2026-09-21
Xiaomi's MiMo team presents CodeMidas, an agentic pipeline that turns implemented functionality in existing codebases into executable RL environments using only source code—no issues or commits needed. Agents explore functionality to write behavioral specs, build tests grounded in the original code's execution, and filter tasks via execution checks and repeated solution rollouts. The dataset spans 5,545 tasks from 3,185 repos across 23 languages and 15 domains. GRPO training on it lifts MiMo-V2.5 by +11.7% on DeepSWE, +17% on ProgramBench and +8.5% on Terminal-Bench v2.1; ablations show more high-quality tasks keep improving performance.
More from coding & agent
- User automates entire video clipping workflow with Nous Research's Hermes agent — Teknium · 2026-11-12
- As models improve, writing good evals gets harder and more important — amankhan · 2026-11-12
- Free AI game jam in SF Nov 14–15 with teams from Anthropic, Google, Meshy, OpenAI — chrisfirst · 2026-11-12
- Omarchy 4.6 lets you share local AI models across your tailnet, says DHH — alexcovo_eth · 2026-11-12
- Coding harnesses hand shell access to inference providers — 4-step gateway mitigation — shensi · 2026-11-12
- Slack's David Hoang: agents beat people on coordination, proposes 'boundary agents' — davidhoang · 2026-11-12