Metraipolis › archive › dc_87d033a52db8

It Takes Workflows to Evolve Better Workflows — arXiv:2610.01026

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.01026
License (as archived)
CC0-1.0
Archived
2026-10-06T12:12:27.133Z
SHA-256 of the archived text
820280acc1f1ad468d7706e19a358e2e6958d41f72c999adfdadcbe961e2fe46
Stable URL of this page
https://metraipolis.com/documento/dc_87d033a52db8

How to cite

It Takes Workflows to Evolve Better Workflows — arXiv:2610.01026 — arXiv (en), https://arxiv.org/abs/2610.01026 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_87d033a52db8 (Metraipolis archive dc_87d033a52db8, sha256 820280acc1f1ad468d7706e19a358e2e6958d41f72c999adfdadcbe961e2fe46)
BibTeX
@misc{dc87d033a52db8,
  title = {It Takes Workflows to Evolve Better Workflows — arXiv:2610.01026},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.01026},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_87d033a52db8; sha256 820280acc1f1ad468d7706e19a358e2e6958d41f72c999adfdadcbe961e2fe46},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.01026. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Tackling complex real-world tasks can exceed the capabilities of a single large language model (LLM), motivating the use of multi-agent workflows that coordinate specialized agents to work together on these tasks. Recent methods train LLMs to construct better workflows from execution outcomes, but they optimize only the workflow generator, while the other agents that build or execute each workflow remain fixed even though every outcome depends on all of them. However, extending training beyond the generator is challenging: the agents are coupled, and a workflow's outcome is a single sparse score that cannot tell which agent causes a failure. We propose FloWright, which leverages the workflow as a harness to optimize workflows. By introducing a hierarchical, structure-aware reward paradigm, FloWright enables one role to self-evolve and two or more roles to co-evolve, with no additional models, labels, or executions. Considering the limitation that workflows are commonly trained and evaluated on data that a single agent can already handle, we further propose DataWright, an adaptive data hardening approach that converts existing datasets into workflow-level tasks with increased difficulty. Across document, slide, chart, code, math, and finance tasks, small open models trained with FloWright achieve improved performance by up to $+7.41\%$, with co-evolving ($+5.03\%$) more roles gaining more than optimizing one of them alone ($+2.83\%$). Our project page: this https URL .

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-10-02 in cs.AI).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.01026 — 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 2026-10-06T10:21:28.146Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-10-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).