Metraipolis › archive › dc_12682c16db39
RSI-Env: An Open-Ended Environment for Recursive Self-Improvement of Agents — arXiv:2611.01736
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.01736
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-03T11:41:38.665Z
- SHA-256 of the archived text
d82a1d0d701b843b71b08512e589339484daf4d2c32f39dd161cd08311e5d04a- Stable URL of this page
https://metraipolis.com/documento/dc_12682c16db39
How to cite
RSI-Env: An Open-Ended Environment for Recursive Self-Improvement of Agents — arXiv:2611.01736 — arXiv (en), https://arxiv.org/abs/2611.01736 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_12682c16db39 (Metraipolis archive dc_12682c16db39, sha256 d82a1d0d701b843b71b08512e589339484daf4d2c32f39dd161cd08311e5d04a)
BibTeX
@misc{dc12682c16db39,
title = {RSI-Env: An Open-Ended Environment for Recursive Self-Improvement of Agents — arXiv:2611.01736},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.01736},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_12682c16db39; sha256 d82a1d0d701b843b71b08512e589339484daf4d2c32f39dd161cd08311e5d04a},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.01736. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Recursive self-improvement (RSI) envisions agent systems that autonomously improve their own models and harnesses, with each updated system continuing to improve itself. Existing approaches typically restrict optimization objectives, editable components, or evaluation procedures, limiting open-ended exploration and continuity across generations. We introduce RSI-Env, a persistent environment for coordinated model-harness improvement, supporting autonomous training, harness modification, and retention of research state. An independent Host Controller manages services, evaluation, and versioning. Under Self, each accepted model-harness version takes over the next improvement round. We define three settings that vary goal specificity and access to benchmark feedback: Open-General, Capability-Directed, and Benchmark-Guided. Across these three settings, we compare Self with two fixed external Teachers on seven tasks, starting from Qwen3.8-27B with a minimal harness. Without benchmark feedback, Open-General Self raises GPQA Diamond accuracy from 65.2% to 73.2% and AIME 2026 accuracy from 53.3% to 56.7%. All three Capability-Directed Self runs achieve target-aligned gains. Knowledge Self demonstrates continued improvement across system updates, with SimpleQA accuracy rising from 82.0% to 85.0% and then 88.0%. Benchmark guidance and Teacher benefits vary by task. These results demonstrate RSI-Env's support for open-ended exploration and sustained recursive improvement within the studied runs, while highlighting challenges in preserving gains across capabilities. Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.CV, nuovi lavori del giorno). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV). Pagina del lavoro: https://arxiv.org/abs/2611.01736 — 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 2026-11-03T11:06:37.508Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).