Metraipolis › archive › dc_37fc0b015eac

Don't Cut Corners: How Training Outside the Prior Makes Simulation-Based Inference More Robust — arXiv:2608.12470

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.12470
License (as archived)
CC0-1.0
Archived
2026-10-19T04:44:10.241Z
SHA-256 of the archived text
40254305b4ae3a586af80e06f3b25837436dff610d7bdf952b2820d4fa131a24
Stable URL of this page
https://metraipolis.com/documento/dc_37fc0b015eac

How to cite

Don't Cut Corners: How Training Outside the Prior Makes Simulation-Based Inference More Robust — arXiv:2608.12470 — arXiv (en), https://arxiv.org/abs/2608.12470 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_37fc0b015eac (Metraipolis archive dc_37fc0b015eac, sha256 40254305b4ae3a586af80e06f3b25837436dff610d7bdf952b2820d4fa131a24)
BibTeX
@misc{dc37fc0b015eac,
  title = {Don't Cut Corners: How Training Outside the Prior Makes Simulation-Based Inference More Robust — arXiv:2608.12470},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.12470},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_37fc0b015eac; sha256 40254305b4ae3a586af80e06f3b25837436dff610d7bdf952b2820d4fa131a24},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.12470. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Large astrophysical simulation campaigns often generate training data by sampling parameters across a Uniform prior box. Due to the proposal's sharp edge, neural posterior estimators struggle to learn accurate approximations near the boundaries. We propose Tailed-Uniform, a family of hybrid proposal distributions for sampling training simulations for robust simulation-based inference. By padding the original hard-truncated training box with decaying tails, Tailed-Uniform-trained networks yield more accurate posteriors near and beyond the edges. We demonstrate these improvements on a family of tail shapes, including a widened Uniform box as a control. Our results suggest that additional simulations near the prior boundary better constrain the networks as it approaches the edge of the training box, even for Uniform assumed priors. We show these advantages on a toy problem and cosmological parameter inference from the matter power spectrum. These benefits increase in high dimensions, where boundaries dominate parameter space volume.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-08-14 in stat.ML).
Categorie arXiv: Instrumentation and Methods for Astrophysics (astro-ph.IM) ; Machine Learning (stat.ML).
Pagina del lavoro: https://arxiv.org/abs/2608.12470 — 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 2026-10-19T04:25:49.381Z dalla pagina https://arxiv.org/catchup/stat.ML/2026-08-14?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).