Metraipolis › archive › dc_976fd009af3e

Stand-In Effects and Pooling Amplification: Design-Based Tests of Whether Biological World Models Use Actions — arXiv:2611.01666

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.01666
License (as archived)
CC0-1.0
Archived
2026-11-03T18:36:58.653Z
SHA-256 of the archived text
1c1d8d2ee1a59cd084ad8f12c9c57616f52acc925676decc9bf08b96983c086f
Stable URL of this page
https://metraipolis.com/documento/dc_976fd009af3e

How to cite

Stand-In Effects and Pooling Amplification: Design-Based Tests of Whether Biological World Models Use Actions — arXiv:2611.01666 — arXiv (en), https://arxiv.org/abs/2611.01666 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_976fd009af3e (Metraipolis archive dc_976fd009af3e, sha256 1c1d8d2ee1a59cd084ad8f12c9c57616f52acc925676decc9bf08b96983c086f)
BibTeX
@misc{dc976fd009af3e,
  title = {Stand-In Effects and Pooling Amplification: Design-Based Tests of Whether Biological World Models Use Actions — arXiv:2611.01666},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.01666},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_976fd009af3e; sha256 1c1d8d2ee1a59cd084ad8f12c9c57616f52acc925676decc9bf08b96983c086f},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.01666. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Action-conditioned world models are increasingly trained on data from randomized experiments, and whether their predictions respond to actions is usually checked by replacing or removing the actions and measuring the loss in accuracy. We show that such diagnostics report action effects where there are none: under autocorrelated designs, future actions stand in for history the model does not see, and a better model learns the stand-in better. When the design law is known, a conditional randomization test written as an e-value is valid in finite samples whatever the model, provided the resampler conditions on the full state the design used; resampling from a simpler law reproduces the false effect. We identify a second failure, specific to world models: when many sites share one action sequence (cells in a microscope field, pixels in a frame), the natural pooled score turns small model errors along the action contrast into overwhelming evidence against a true effect. In a pixel world, its median log e-value fell from $-17.3$ to $-6832$ as frames grew from 64 to 4096 pixels, while the oracle's rose. Tempering the score by a design effect estimated on training data changes the bet, not the validity. Our own preregistered analysis of optogenetic ERK experiments failed in exactly this way; a tempered statistic, selected on synthetic data and registered before the data were obtained, then rejected the null on a never-opened field ($E\approx 1.1\times10^{10}$). Every analysis is preregistered, fixed in a dated protocol before it was run, or labelled exploratory, and every reported result is generated from hashed result files.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG) ; Quantitative Methods (q-bio.QM); Methodology (stat.ME).
Pagina del lavoro: https://arxiv.org/abs/2611.01666 — 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 2026-11-03T18:12:54.514Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).