Metraipolis › archive › dc_96e4f3f2cc86
AnyAct: Universal Action for Self-Evolving Agents — arXiv:2609.37025
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.37025
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T16:23:24.154Z
- SHA-256 of the archived text
085dc6f195a4044c93580f7f503821f72f88c85306363dbfffeffb973e9e3ac8- Stable URL of this page
https://metraipolis.com/documento/dc_96e4f3f2cc86
How to cite
AnyAct: Universal Action for Self-Evolving Agents — arXiv:2609.37025 — arXiv (en), https://arxiv.org/abs/2609.37025 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_96e4f3f2cc86 (Metraipolis archive dc_96e4f3f2cc86, sha256 085dc6f195a4044c93580f7f503821f72f88c85306363dbfffeffb973e9e3ac8)
BibTeX
@misc{dc96e4f3f2cc86,
title = {AnyAct: Universal Action for Self-Evolving Agents — arXiv:2609.37025},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.37025},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_96e4f3f2cc86; sha256 085dc6f195a4044c93580f7f503821f72f88c85306363dbfffeffb973e9e3ac8},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.37025. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
As large language models (LLMs) advance, AI agents are increasingly deployed in open-world environments to tackle complex sequential tasks (e.g., document processing, cross-application collaboration), relying heavily on actions ranging from GUI operations to semantic APIs. However, three core challenges persist: the "scale dilemma" of massive tool ecosystems exceeding LLM context windows, the "non-stationarity" of tool quality due to updates or outages, and the "heterogeneity" of feedback formats (pixels, text, structured data) creating information silos. To address these, we propose AnyAct, a universal action layer that unifies available capabilities into a self-evolving action space, enabling agents to operate efficiently and reliably in large-scale, dynamic tool ecosystems. AnyAct's core design focuses on two objectives: constructing this action space via hierarchical progressive retrieval (filtering task-relevant actions) and test-time reliability evolution (pruning unreliable actions), and enabling reliability-aware action orchestration through a heterogeneous observation grounding module that unifies multi-modal feedback. Additionally, it defines a hybrid action space (primitive + semantic actions) and optimizes for a balance between task success rate and execution cost. Evaluations on LiveMCPBench and OSMCP (a new benchmark we developed for multi-granularity action collaboration) demonstrate state-of-the-art performance. AnyAct delivers substantial performance gains over baseline methods across various LLM base models on LiveMCPBench and improvements are particularly notable for models with constrained native capabilities. On OSMCP, it achieves 77.27% overall success with only 50 steps, which is half the steps required by most competitors. 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/2609.37025 — 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 2026-10-07T15:54:53.579Z 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).