Metraipolis › archive › dc_b60d940887cf

WEFT: Scaling Tool-Use Post-Training for General-Purpose Agents — arXiv:2609.36887

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.36887
License (as archived)
CC0-1.0
Archived
2026-10-07T16:23:11.435Z
SHA-256 of the archived text
8c086f6923df08c568e746c1c596f1a61366c6c6ee43b31363a54e3d561b7d54
Stable URL of this page
https://metraipolis.com/documento/dc_b60d940887cf

How to cite

WEFT: Scaling Tool-Use Post-Training for General-Purpose Agents — arXiv:2609.36887 — arXiv (en), https://arxiv.org/abs/2609.36887 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_b60d940887cf (Metraipolis archive dc_b60d940887cf, sha256 8c086f6923df08c568e746c1c596f1a61366c6c6ee43b31363a54e3d561b7d54)
BibTeX
@misc{dcb60d940887cf,
  title = {WEFT: Scaling Tool-Use Post-Training for General-Purpose Agents — arXiv:2609.36887},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.36887},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_b60d940887cf; sha256 8c086f6923df08c568e746c1c596f1a61366c6c6ee43b31363a54e3d561b7d54},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.36887. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Recent efforts to scale tool-use post-training have largely centered on the synthesis of executable environments, which constitute only one component of a broader agentic interaction system comprising the environment, task, agent harness, and evaluator. Scaling environments in isolation, however, does not guarantee commensurate gains in model performance, because reliable learning signals depend on coherent interactions among all components of the agentic interaction system. To address this problem, we introduce WEFT (Whole-system Evolution For Tool-use Post-training), which couples scalable agentic interaction system construction, execution-driven self-evolution, and stable post-training. WEFT scales agentic interaction system construction across environment breadth, task complexity, and interaction diversity. Execution-driven self-evolution iteratively uses execution traces and state evidence to attribute failures and revise the responsible components, with fresh rollouts evaluating the changes and providing evidence for subsequent evolution rounds. For stable post-training at scale, WEFT addresses both optimization and execution reliability: prefix-preserving sampling retains verified progress and atomic-turn credit assignment localizes learning signals, while MegaMCP maintains isolated, recoverable state across concurrent rollouts over shared tool services. Extensive experiments across various models and benchmarks demonstrate the effectiveness of WEFT for tool-use post-training. WEFT-8B and WEFT-14B outperform all evaluated matched-size environment-scaling baselines on BFCL V4, $\tau^2$-Bench, and Claw-Eval. In particular, WEFT-14B improves over Agent-World-14B by 6.41, 2.23, and 12.27 percentage points. WEFT-35B-A3B further extends these gains to more challenging long-horizon workflow benchmarks, including Toolathlon-Verified and AutomationBench.

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/2609.36887 — 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 2026-10-07T15:54:53.557Z 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).