Metraipolis › archive › dc_87fce47503a7

Harness-Zero: Harness Distillation via Agent-as-Harness — arXiv:2609.24974

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.24974
License (as archived): CC0-1.0
Archived: 2026-10-09T06:50:45.316Z
SHA-256 of the archived text: 54fc1e658c05d1936e665d558443b1e85ed2b5b270f8544a4db233a16ea5f1f0
Stable URL of this page: https://metraipolis.com/documento/dc_87fce47503a7

How to cite

Harness-Zero: Harness Distillation via Agent-as-Harness — arXiv:2609.24974 — arXiv (en), https://arxiv.org/abs/2609.24974 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_87fce47503a7 (Metraipolis archive dc_87fce47503a7, sha256 54fc1e658c05d1936e665d558443b1e85ed2b5b270f8544a4db233a16ea5f1f0)

BibTeX

@misc{dc87fce47503a7,
  title = {Harness-Zero: Harness Distillation via Agent-as-Harness — arXiv:2609.24974},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.24974},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_87fce47503a7; sha256 54fc1e658c05d1936e665d558443b1e85ed2b5b270f8544a4db233a16ea5f1f0},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.24974. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Agent harnesses, the external systems that mediate model-environment interaction, can substantially improve agent performance, but their gains remain tied to the harness at deployment. Because the best harness varies across domains, instances, and models, a general-purpose agent must either settle for a suboptimal shared harness or route among an ever-growing set of specialized ones. We therefore study agent harness distillation: using a domain- or instance-optimized harness as training-time guidance and transferring the behaviors it induces into model weights, so that its gains survive under a single fixed target harness. The challenge is that the two harnesses differ in action space and available information, so guidance from the optimized harness cannot serve directly as supervision for the target one. We introduce Harness-Zero, which enables harness distillation through agent-as-harness. Guided by the optimized harness, a harnessing agent corrects student responses before execution in the target harness's action space, turning harness guidance into training demonstrations. Fine-tuning on the resulting trajectories internalizes harness-induced behavior into the model, so the specialized harness can be removed at deployment. Our experiments spanning knowledge work, tool use, and science domains show that: (1) For frontier LLMs using the same evolved harness, agent-as-harness outperforms code-as-harness. (2) With the specialized harness removed at deployment, Harness-Zero improves the base model's macro-average task success from 23.3% to 44.3%, even exceeding the 41.7% it reaches with that harness still attached. (3) Harness-Zero recovers harness-induced behaviors absent from the base model, with 82.3% average recovery across 28 patterns in the three domains.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-22 in cs.CL).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Computation and Language (cs.CL); Neural and Evolutionary Computing (cs.NE).
Pagina del lavoro: https://arxiv.org/abs/2609.24974 — 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 2026-10-09T06:48:45.159Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-22?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).