Metraipolis › archive › dc_336ef4c72ca2

DSReg: Provably Recovering Individual World Latents without Reconstruction — arXiv:2610.09457

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.09457
License (as archived)
CC0-1.0
Archived
2026-10-08T11:11:02.366Z
SHA-256 of the archived text
74339fd8584454fcf0d59021fa65f6f3a46cbe69300826843c3ff8a8597b232e
Stable URL of this page
https://metraipolis.com/documento/dc_336ef4c72ca2

How to cite

DSReg: Provably Recovering Individual World Latents without Reconstruction — arXiv:2610.09457 — arXiv (en), https://arxiv.org/abs/2610.09457 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_336ef4c72ca2 (Metraipolis archive dc_336ef4c72ca2, sha256 74339fd8584454fcf0d59021fa65f6f3a46cbe69300826843c3ff8a8597b232e)
BibTeX
@misc{dc336ef4c72ca2,
  title = {DSReg: Provably Recovering Individual World Latents without Reconstruction — arXiv:2610.09457},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.09457},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_336ef4c72ca2; sha256 74339fd8584454fcf0d59021fa65f6f3a46cbe69300826843c3ff8a8597b232e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.09457. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Methods that recover individual latent variables of the world, from nonlinear ICA to dictionary learning and causal representation learning, anchor the latents to observations through reconstruction, auxiliary supervision, or distributional asymmetries such as non-Gaussianity. Methods without these anchors, including joint-embedding predictive architectures (JEPAs), identify the latent state only up to a linear transformation, so individual latents remain mixed. We close this gap: individual world latents can be provably recovered with no reconstruction, no decoder, and no labels. The key condition is Structural Diversity: different latents leave distinct dependency footprints on observations, just as no two snowflakes are alike. Building on the linear identifiability that LeJEPA provides, we prove that under Structural Diversity, DSReg (Dependency-Sparsity Regularization) recovers individual world latents up to signed permutation, without reconstruction or a decoder. It applies post hoc to any linearly identified representation, reusing trained checkpoints at no loss over joint training, and establishes the first fully identifiable JEPA that recovers every world latent. Moreover, as a condition on dependency footprints, Structural Diversity is strictly weaker than all structural conditions of prior identifiable latent variable models. Across synthetic regimes, world model probes, learned visual encoders, and external renderers, DSReg preserves dense prediction while improving individual-latent recovery and downstream use with scales.

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); Robotics (cs.RO); Machine Learning (stat.ML).
Pagina del lavoro: https://arxiv.org/abs/2610.09457 — 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 2026-10-08T10:14:06.089Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).