Metraipolis › archive › dc_63f2137ea005

LegoEnv: Self-Evolving Agentic Environments via Modular Recombination — arXiv:2611.01834

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.01834
License (as archived)
CC0-1.0
Archived
2026-11-03T18:18:12.582Z
SHA-256 of the archived text
3c9431010dd4b24648faf7200cd3a5eb06fdb047a83e9076ee33d54c02f928ed
Stable URL of this page
https://metraipolis.com/documento/dc_63f2137ea005

How to cite

LegoEnv: Self-Evolving Agentic Environments via Modular Recombination — arXiv:2611.01834 — arXiv (en), https://arxiv.org/abs/2611.01834 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_63f2137ea005 (Metraipolis archive dc_63f2137ea005, sha256 3c9431010dd4b24648faf7200cd3a5eb06fdb047a83e9076ee33d54c02f928ed)
BibTeX
@misc{dc63f2137ea005,
  title = {LegoEnv: Self-Evolving Agentic Environments via Modular Recombination — arXiv:2611.01834},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.01834},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_63f2137ea005; sha256 3c9431010dd4b24648faf7200cd3a5eb06fdb047a83e9076ee33d54c02f928ed},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.01834. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

LLM agents learn by interacting with environments, yet existing environments are designed for specific tasks and fail to generalize to new ones. Recent methods automate environment construction with LLMs. However, they generate each environment from scratch, neither reusing parts shared across environments nor carrying a fix from one environment to the rest. To address this, we propose a new paradigm of evolving environment atoms, which composes environments from reusable atoms (intent, rule, data and server) and evolves the atoms when an environment fails to give reliable feedback (e.g., an application lacks a tool the task requires). To instantiate this paradigm, we introduce LegoEnv, which builds 5,465 verified environments over 223 applications, and one round of evolution raises the share of solvable environments from 71.6% to 88.3%. On AutomationBench, SFT on these environments lifts Qwen3.6-27B from 16.0% to 32.0%, and RL with their verifiers lifts Qwen3.8-27B from 37.6% to 40.3%. Further analyses show steady gains with corpus size, transfer to unseen applications, and failures concentrated in a few shared root causes.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2611.01834 — 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 2026-11-03T18:09:53.629Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).