Metraipolis › archive › dc_770957cc5a30

KPHarness: Co-Evolving Knowledge and Probing for Long-Horizon Gaming Harnesses — arXiv:2610.27417

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.27417
License (as archived)
CC0-1.0
Archived
2026-10-21T07:06:33.309Z
SHA-256 of the archived text
9017378b2d928b9511e2635720d53546d7f06262d827a24a7a793b98bbf6dce8
Stable URL of this page
https://metraipolis.com/documento/dc_770957cc5a30

How to cite

KPHarness: Co-Evolving Knowledge and Probing for Long-Horizon Gaming Harnesses — arXiv:2610.27417 — arXiv (en), https://arxiv.org/abs/2610.27417 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_770957cc5a30 (Metraipolis archive dc_770957cc5a30, sha256 9017378b2d928b9511e2635720d53546d7f06262d827a24a7a793b98bbf6dce8)
BibTeX
@misc{dc770957cc5a30,
  title = {KPHarness: Co-Evolving Knowledge and Probing for Long-Horizon Gaming Harnesses — arXiv:2610.27417},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.27417},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_770957cc5a30; sha256 9017378b2d928b9511e2635720d53546d7f06262d827a24a7a793b98bbf6dce8},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.27417. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

A failed gaming trajectory reveals that an agent needs intervention, but not necessarily how its execution harness should change. Effective harness adaptation therefore requires gathering diagnostic evidence, rather than merely proposing alternative recovery rules. We present KPHarness, a framework that adapts the external execution harness of a frozen vision-language-action model through knowledge-probing co-evolution. Recovery knowledge supplies context-specific intervention hypotheses that guide active probing; probe outcomes reveal when these hypotheses apply and provide evidence for refining both recovery knowledge and later probe selection. This reciprocal process allows the harness to improve not only which interventions it applies, but also how it investigates failures. KPHarness separates fast policy execution from a slow research loop that translates accumulated knowledge and probing evidence into bounded harness revisions. Each revision is compared with the incumbent under matched sandbox conditions and accepted only when the paired evaluation shows improvement. Recoveries supported by further validation are retained for cross-task reuse, while immediate interaction feedback and delayed episode-success credit update action- and knowledge-probing policies during adaptation. Experiments on Minecraft tasks spanning mining, combat, and crafting show improved task coverage over bare policy execution, a manually specified fixed harness, and knowledge-only harness evolution with non-adaptive probing. Evaluation on previously unseen tasks further shows improved task completion with the evolved harness, knowledge bank, and probing policies held fixed,

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2610.27417 — 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 2026-10-21T07:05:43.408Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).