Metraipolis › archive › dc_a262f1df8fe7
AREX-2: Advancing Self-Improving Agents through Long-Horizon Reflective Tasks — arXiv:2609.38288
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.38288
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T12:41:40.800Z
- SHA-256 of the archived text
8650785bc903addb0cb23d305a0739109fd24b5846b1ef15f159aa7c94c8128e- Stable URL of this page
https://metraipolis.com/documento/dc_a262f1df8fe7
How to cite
AREX-2: Advancing Self-Improving Agents through Long-Horizon Reflective Tasks — arXiv:2609.38288 — arXiv (en), https://arxiv.org/abs/2609.38288 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_a262f1df8fe7 (Metraipolis archive dc_a262f1df8fe7, sha256 8650785bc903addb0cb23d305a0739109fd24b5846b1ef15f159aa7c94c8128e)
BibTeX
@misc{dca262f1df8fe7,
title = {AREX-2: Advancing Self-Improving Agents through Long-Horizon Reflective Tasks — arXiv:2609.38288},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.38288},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_a262f1df8fe7; sha256 8650785bc903addb0cb23d305a0739109fd24b5846b1ef15f159aa7c94c8128e},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.38288. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
We present AREX-2, an effort to advance the self-improving capability of LLM agents, which we define as the ability to iteratively refine a solution at test time. This ability rests on two complementary capabilities: reflection, which produces a solution better than the current one, and long-horizon execution, which keeps the iteration effective over many rounds. We hypothesize that both capabilities are domain-agnostic, and can therefore be learned in scenarios that are well suited for supervision. Accordingly, we synthesize long-horizon improvement trajectories from machine learning and algorithmic programming tasks, two domains that offer verifiable feedback and reward sustained iteration. Trained on this data, our agent, built on Qwen3.8-27B, achieves strong results on MLE-bench Lite (81.8) and Frontier-CS (70.7), transfers to deep research with 84.0 on BrowseComp, 52.6 on HLE, 92.2 on GAIA, and 93.8 on DeepSearchQA, and keeps improving as its budget of rounds grows. These results show that long-horizon reflective data is an effective route toward self-improving agents. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-10-01 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2609.38288 — 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 2026-10-07T12:14:13.606Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-10-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).