Metraipolis › archive › dc_bae426eb4e9d

Predictive gains, causal losses: why interventional competence cannot be learned from data alone — arXiv:2611.03861

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.03861
License (as archived)
CC0-1.0
Archived
2026-11-04T18:33:43.849Z
SHA-256 of the archived text
ed327338bdab588a6d874c73a696142f423cddf1b4d111290e9c087721308232
Stable URL of this page
https://metraipolis.com/documento/dc_bae426eb4e9d

How to cite

Predictive gains, causal losses: why interventional competence cannot be learned from data alone — arXiv:2611.03861 — arXiv (en), https://arxiv.org/abs/2611.03861 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_bae426eb4e9d (Metraipolis archive dc_bae426eb4e9d, sha256 ed327338bdab588a6d874c73a696142f423cddf1b4d111290e9c087721308232)
BibTeX
@misc{dcbae426eb4e9d,
  title = {Predictive gains, causal losses: why interventional competence cannot be learned from data alone — arXiv:2611.03861},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.03861},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_bae426eb4e9d; sha256 ed327338bdab588a6d874c73a696142f423cddf1b4d111290e9c087721308232},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.03861. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Retaining predictive gains can require persistent errors in a model's response to interventions. For deterministic predictors with finite second moments under squared-error loss, any observational advantage over the best predictor using only an outcome's direct causes requires dependence on other inputs, regardless of the learning rule. When admissible interventions vary inputs while preserving the worldly outcome, this dependence yields a positive lower bound on root-mean-square intervention-response error. We identify conditions under which population gradient updates introduce this dependence from a parent-only baseline or damage a previously correct mechanism through shared parameters while improving other predictions. These dependencies can also make model behavior vary with how the same intervention is realized. We characterize interventional alignment through a common output mapping across internal states and physical processes realizing the same intervention target and relevant context. Correction is limited by the causal hypotheses currently admitted within inquiry. The relevant hypothesis class comprises these causal alternatives, rather than the function class the computational system could in principle represent. For predictions restricted to these alternatives, interventional mean error cannot fall below the best approximation they permit, whether evidence is observational or interventional. Reducing error below a positive limit requires admitting better hypotheses, which may entail revising the causal variables, mechanism structure, or intervention interpretation. Interventional criticism constrains replacements without determining them. We conjecture that reliably acquiring and maintaining interventional competence when causal commitments fail requires autonomous participation in proposing alternatives, testing their consequences, and revising them in response to error.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2611.03861 — 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 2026-11-04T18:32:37.249Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).