Metraipolis › archive › dc_b1e20dddbd78

Good Views Are Not Enough: Global Counterexamples for Gaussian-Regularized JEPAs — arXiv:2611.12137

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.12137
License (as archived)
CC0-1.0
Archived
2026-11-10T18:43:55.755Z
SHA-256 of the archived text
6fe82132b5a091a2c0b86071fe0d661daed93fbb98ab0a41c848a8a5316ea0a9
Stable URL of this page
https://metraipolis.com/documento/dc_b1e20dddbd78

How to cite

Good Views Are Not Enough: Global Counterexamples for Gaussian-Regularized JEPAs — arXiv:2611.12137 — arXiv (en), https://arxiv.org/abs/2611.12137 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_b1e20dddbd78 (Metraipolis archive dc_b1e20dddbd78, sha256 6fe82132b5a091a2c0b86071fe0d661daed93fbb98ab0a41c848a8a5316ea0a9)
BibTeX
@misc{dcb1e20dddbd78,
  title = {Good Views Are Not Enough: Global Counterexamples for Gaussian-Regularized JEPAs — arXiv:2611.12137},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.12137},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_b1e20dddbd78; sha256 6fe82132b5a091a2c0b86071fe0d661daed93fbb98ab0a41c848a8a5316ea0a9},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.12137. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Good views do not suffice to guarantee scene invariance or uniqueness up to rotation at the global optima of Gaussian-regularized JEPAs. We give globally certified counterexamples showing that the uniqueness premise of the Gaussian extension in Good Views fails for specified soft objectives. Gaussian marginal matching can reward retaining independently resampled observation variation, and this benefit can exceed its prediction cost. In a four-state model, comparison against the entire scene-only class proves that every global minimizer splits a scene and admits an equal-value partner outside its rotation orbit. In a 512-dimensional model with a complete scene basis, a spectral gap, and exact population whitening, pooled Epps-Pulley regularization still forces every global minimizer to retain nuisance and yields positive cross-modal alignment error even after the best rotation when modality nuisances are independent. Better scene alignment can also accompany worse prediction. With matched 512-label readout budgets and 1% observation noise, whitened prediction JEPA aligns better to the scene than next-token prediction but incurs 0.2872 nats/token higher linear-readout test NLL. A Bayes-risk decomposition distinguishes information lost through the observation channel from the fitted readout's difficulty using retained information. These results connect three determinants of learned representations: what views share, what regularization favors, and what a prediction task can use.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2611.12137 — 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 2026-11-10T18:37:37.678Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).