Metraipolis › archive › dc_c375a8a67a20

Recursive self-improvement of AI research agents — arXiv:2609.26457

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.26457
License (as archived): CC0-1.0
Archived: 2026-10-08T13:45:46.815Z
SHA-256 of the archived text: 9129e356cb2c5e27a5d5aab472e811fa60d3498202a1fd0cfb2893c5d49190ba
Stable URL of this page: https://metraipolis.com/documento/dc_c375a8a67a20

How to cite

Recursive self-improvement of AI research agents — arXiv:2609.26457 — arXiv (en), https://arxiv.org/abs/2609.26457 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_c375a8a67a20 (Metraipolis archive dc_c375a8a67a20, sha256 9129e356cb2c5e27a5d5aab472e811fa60d3498202a1fd0cfb2893c5d49190ba)

BibTeX

@misc{dcc375a8a67a20,
  title = {Recursive self-improvement of AI research agents — arXiv:2609.26457},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.26457},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_c375a8a67a20; sha256 9129e356cb2c5e27a5d5aab472e811fa60d3498202a1fd0cfb2893c5d49190ba},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.26457. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

AI agents are beginning to automate research and development across the AI stack, from improving training efficiency to optimizing inference. A natural next step is to improve the research efficiency of the agents themselves. When an AI research agent's own code is the object of optimization, each accepted rewrite becomes the agent that the next round edits. We refer to this loop as recursive self-improvement. Its significance lies in a long-standing trend, in which increased cumulative spending on R&D yields diminishing returns. Sustained self-improvement offers a way to counter this trend. We present AIDE^2, a system that implements this loop for a frontier AI research agent. It proposes changes to its own code, benchmarks modified versions of itself on a suite of AI R&D tasks, and keeps the changes that perform best on hidden evaluations. In an autonomous 8-day run, AIDE^2 discovered seven successive improvements, ranging from a new search policy to memory mechanisms that compress and manage the agent's growing context. These gains generalize to four held-out benchmarks spanning machine learning engineering, heuristic algorithm engineering, and physics-based weather forecasting, the last of which is out of distribution from the selection tasks. On all four, the strongest discovered agent matches or exceeds a human-engineered production research agent that ranks among the strongest on FML-Bench. On a separate held-out task family, the discovered agents also exhibit reduced reward hacking, a property the loop never explicitly optimized for: the rate falls from 55% to 32% during the run, 7 percentage points below the human-engineered agent. Together, these results show that an AI research agent can improve its own research efficiency through recursive self-improvement, and that these gains transfer to tasks and domains the loop never encountered.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-23 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Machine Learning (cs.LG); Software Engineering (cs.SE).
Pagina del lavoro: https://arxiv.org/abs/2609.26457 — 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 2026-10-08T13:18:38.169Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-23?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).