Metraipolis › archive › dc_ebc0bc74a531
Training LLM Agents to Automate LLM Post-Training with RSI-Dojo — arXiv:2611.03698
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.03698
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-04T21:36:25.249Z
- SHA-256 of the archived text
04bd28aff51b95a494939fc65244ce75685d8cc420e2cba8ed44f0dc2549e505- Stable URL of this page
https://metraipolis.com/documento/dc_ebc0bc74a531
How to cite
Training LLM Agents to Automate LLM Post-Training with RSI-Dojo — arXiv:2611.03698 — arXiv (en), https://arxiv.org/abs/2611.03698 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ebc0bc74a531 (Metraipolis archive dc_ebc0bc74a531, sha256 04bd28aff51b95a494939fc65244ce75685d8cc420e2cba8ed44f0dc2549e505)
BibTeX
@misc{dcebc0bc74a531,
title = {Training LLM Agents to Automate LLM Post-Training with RSI-Dojo — arXiv:2611.03698},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.03698},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ebc0bc74a531; sha256 04bd28aff51b95a494939fc65244ce75685d8cc420e2cba8ed44f0dc2549e505},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.03698. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Recursive Self-Improvement (RSI) aims to enable AI systems to autonomously build and improve AI systems themselves, including models, harnesses, and underlying infrastructure. Within this broader agenda, a central capability is enabling agents to train and improve base models. However, existing studies on automated model training primarily focused on workflow- or harness-level optimizations, rather than internalizing the ability to conduct post-training into the model itself. Developing this capability requires executable model training environments that can produce high-quality trajectories demonstrating how to select data, train models, diagnose failures, and iteratively improve results. In this paper, we introduce RSI-Dojo, a collection of executable model training environments designed to teach LLM agents to conduct post-training. To construct RSI-Dojo, we develop RSI-Craft, an automatic pipeline that converts human research artifacts into executable model training environments. RSI-Craft automatically collects research papers from top venues, instantiates datasets and metrics into a unified interface, and reproduces evaluation dependencies. It then filters tasks by data size, execution reliability, and discriminability, yielding RSI-Dojo with 281 seed model training tasks. Building on RSI-Dojo, we employ a trajectory collection method with an improved harness to collect high-quality model-training trajectories from frontier models, then retain examples through rubric-based and outcome-based filtering. We fine-tune Qwen3.5-9B and Qwen3.6-35B-A3B on these trajectories. The resulting RSI-Dojo-9B and RSI-Dojo-35B-A3B improve over their base models by 8.74 and 16.87 PostTrainBench points, respectively. RSI-Dojo-35B-A3B reaches 29.95 on PostTrainBench and 53.29 on PostTrainBench-Lite, matching the performance of frontier models such as GLM-5.2 and GPT 5.5. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2611.03698 — 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 2026-11-04T21:35:49.485Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).