Metraipolis › archive › dc_b0271dfd6eee

EASE: Behavior-Adaptive Skill Curation for Self-Evolving Agents — arXiv:2609.36746

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.36746
License (as archived): CC0-1.0
Archived: 2026-10-07T16:22:58.379Z
SHA-256 of the archived text: d119ff040769290ee9871f458cf84d792257508e0328dde6d0c72d11488973d2
Stable URL of this page: https://metraipolis.com/documento/dc_b0271dfd6eee

How to cite

EASE: Behavior-Adaptive Skill Curation for Self-Evolving Agents — arXiv:2609.36746 — arXiv (en), https://arxiv.org/abs/2609.36746 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_b0271dfd6eee (Metraipolis archive dc_b0271dfd6eee, sha256 d119ff040769290ee9871f458cf84d792257508e0328dde6d0c72d11488973d2)

BibTeX

@misc{dcb0271dfd6eee,
  title = {EASE: Behavior-Adaptive Skill Curation for Self-Evolving Agents — arXiv:2609.36746},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.36746},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_b0271dfd6eee; sha256 d119ff040769290ee9871f458cf84d792257508e0328dde6d0c72d11488973d2},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.36746. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Agent skills provide a lightweight mechanism for self-evolving agents to accumulate reusable procedural knowledge without updating model parameters. However, existing learned skill curators typically optimize curation without explicitly modeling downstream executor behavior. We show that this can cause systematic cross-executor degradation: curators trained with different executors perform best when paired with their own training executor, indicating that effective skill curation is executor-dependent. We formulate behavior-adaptive skill curation and introduce EASE, a framework that learns a single curator that adapts its decisions to different executor behaviors. EASE maintains an online behavioral profile of recent execution patterns and conditions the curator on this profile, the current trajectory, and retrieved skills to add, modify, or remove skills from an evolving repository. We train the shared curator jointly across multiple frozen executors with reinforcement learning, using retrieval-aware and behavior-aware temporal attribution to focus optimization on curation actions with observable downstream influence. Across ALFWorld, ScienceWorld, and WebShop, with executors ranging from Qwen3-8B/32B and GPT-OSS-120B to unseen Kimi K2.6, DeepSeek V4 Flash, and Gemini 3.5 Flash, EASE outperforms strong skill- and memory-based baselines without per-executor finetuning. EASE also maintains 34.5--41.0% fewer skills, improves skill retrieval by 36.3--38.7% and measured edit utility by 51.8--60.0%, and reduces deployment-time inference tokens by 9.1--14.5%. These results establish behavior-adaptive skill curation as an effective principle for building self-evolving agents.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-30 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.36746 — 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 2026-10-07T15:54:53.535Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-30?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).