Metraipolis › archive › dc_647bdbc68d8a

EvolveNet: Collaborative Harness Evolution for Agent Self-Improvement — arXiv:2608.04968

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.04968
License (as archived)
CC0-1.0
Archived
2026-10-31T11:35:25.254Z
SHA-256 of the archived text
36a670061e28e80e361f320d934b2a1ea5bd26aa3a11debb13ccc031ae82a395
Stable URL of this page
https://metraipolis.com/documento/dc_647bdbc68d8a

How to cite

EvolveNet: Collaborative Harness Evolution for Agent Self-Improvement — arXiv:2608.04968 — arXiv (en), https://arxiv.org/abs/2608.04968 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_647bdbc68d8a (Metraipolis archive dc_647bdbc68d8a, sha256 36a670061e28e80e361f320d934b2a1ea5bd26aa3a11debb13ccc031ae82a395)
BibTeX
@misc{dc647bdbc68d8a,
  title = {EvolveNet: Collaborative Harness Evolution for Agent Self-Improvement — arXiv:2608.04968},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.04968},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_647bdbc68d8a; sha256 36a670061e28e80e361f320d934b2a1ea5bd26aa3a11debb13ccc031ae82a395},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.04968. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

An LLM agent can acquire persistent improvements by evolving its harness---the executable program governing model calls, tool use, verification, and recovery---without updating model weights. When the same harness serves users with different workloads, isolated evolution limits the sharing of useful discoveries, while centralized adaptation can introduce conflicting task-specific changes. We introduce EvolveNet, a framework for collaborative harness evolution that connects local specialization with shared capability accumulation. Clients evolve the same broadcast harness on their local workloads in parallel, and a server composes their adaptations into an updated shared program. Redistribution makes accepted discoveries available to all clients, allowing subsequent local evolution to build on improvements accumulated across the collaboration. EvolveNet supports this process through evidence-guided, scope-typed program aggregation and behavioral validation. Our theoretical analysis provides sufficient conditions for retaining and combining complementary improvements despite program interactions. Experiments across five settings show improved shared-harness performance, with ablation and behavioral analyses supporting the contribution of composition beyond specialist selection alone.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-06 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2608.04968 — 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 2026-10-31T11:24:48.804Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-08-06?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).