Metraipolis › archive › dc_43e338398cfe
CodeMidas: Scaling Agentic Coding RL Environments from Code Itself — arXiv:2609.22068
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.22068
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-10T01:05:38.930Z
- SHA-256 of the archived text
eba10b961c75c41941fc3dd20ed91d3b87ebb4311baffe6a969bedcecd31db5c- Stable URL of this page
https://metraipolis.com/documento/dc_43e338398cfe
How to cite
CodeMidas: Scaling Agentic Coding RL Environments from Code Itself — arXiv:2609.22068 — arXiv (en), https://arxiv.org/abs/2609.22068 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_43e338398cfe (Metraipolis archive dc_43e338398cfe, sha256 eba10b961c75c41941fc3dd20ed91d3b87ebb4311baffe6a969bedcecd31db5c)
BibTeX
@misc{dc43e338398cfe,
title = {CodeMidas: Scaling Agentic Coding RL Environments from Code Itself — arXiv:2609.22068},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.22068},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_43e338398cfe; sha256 eba10b961c75c41941fc3dd20ed91d3b87ebb4311baffe6a969bedcecd31db5c},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.22068. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Training capable coding agents via reinforcement learning (RL) requires diverse tasks with reliable verifiers. Open-source codebases offer a rich source of such tasks, while existing methods typically rely on development artifacts such as issues and commits, limiting the range of tasks that can be extracted. To better scale RL environments, we present CodeMidas, an agentic pipeline that turns implemented functionality in existing codebases into executable RL environments using source code as its only task-specific input. CodeMidas allocates agentic compute to every stage of environment construction: agents explore implemented functionality to formulate behavioral specifications, construct tests grounded in execution of the original code, and validate and filter candidate tasks through execution checks and repeated solution rollouts. The resulting dataset has 5,545 training tasks from 3,185 open-source codebases spanning 23 programming languages and 15 technical domains. Training MiMo-V2.5 on these tasks with GRPO improves performance on all five diverse benchmarks, covering issue repair (DeepSWE + 11.7%), whole-program construction (ProgramBench +17%), and terminal work (Terminal-Bench v2.1 +8.5%). Ablations show that increasing the number of high-quality training tasks improves performance. Trajectory analysis shows the RL-trained agent demonstrates better behaviors like increasing codebase exploration and more diverse self-verification. These results establish source code as a scalable foundation for constructing RL environments that improve coding agents across diverse software tasks. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-21 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2609.22068 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-10T00:23:42.485Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-21?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).