Metraipolis › archive › dc_5029773eb4ac

Constraint Tree Exploration for Learning from Language Feedback — arXiv:2610.09107

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.09107
License (as archived)
CC0-1.0
Archived
2026-10-08T11:08:55.115Z
SHA-256 of the archived text
f9ba7bbeb3d74d83c7526102303befb6551f3217db5ccc1e5e45a66d2baaaede
Stable URL of this page
https://metraipolis.com/documento/dc_5029773eb4ac

How to cite

Constraint Tree Exploration for Learning from Language Feedback — arXiv:2610.09107 — arXiv (en), https://arxiv.org/abs/2610.09107 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_5029773eb4ac (Metraipolis archive dc_5029773eb4ac, sha256 f9ba7bbeb3d74d83c7526102303befb6551f3217db5ccc1e5e45a66d2baaaede)
BibTeX
@misc{dc5029773eb4ac,
  title = {Constraint Tree Exploration for Learning from Language Feedback — arXiv:2610.09107},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.09107},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_5029773eb4ac; sha256 f9ba7bbeb3d74d83c7526102303befb6551f3217db5ccc1e5e45a66d2baaaede},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.09107. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Natural-language feedback in interactive learning often explains why an action failed by pointing to violated requirements. Misinterpreting this feedback can lead an agent to rule out valid solutions. We study this setting by modeling user intent as latent constraints over an action space and formulating learning from language feedback as pure exploration over feasible regions. We introduce TRACE, an algorithm that organizes candidate constraints in a tree and tests each proposed refinement by generating actions that satisfy it. TRACE commits to the refinement only if the resulting feedback does not contradict it over repeated tests. We distinguish two ways of using the same feedback: (i) falsification, which detects contradictions to the constraint set currently being tested, and (ii) identification, which may additionally name a violated constraint. We prove high-probability coverage bounds with dependence on the candidate class size $H$ for TRACE-Falsification. With reliable identification, TRACE-Identification can replace this dependence by $K/p_{\mathrm{ext}}$, where $K$ is the number of latent constraints and $p_{\mathrm{ext}}$ lower-bounds the probability of extracting a missing true constraint from informative feedback. We evaluate TRACE across six language-feedback tasks. On RecMovie, TRACE-Identification achieves 73% and 86% final-output success under caps of 20 and 60 evaluated outputs, compared with at most 42% and 48% for the evaluated prompting baselines given the same feedback and output caps. Controlled identity-corruption experiments further show greater robustness than direct accumulation when the falsification detector remains reliable.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2610.09107 — 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 2026-10-08T10:14:05.901Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).