Metraipolis › archive › dc_578fc183bbcc

Beyond Invariance: Factor-wise Selective Treatment of Shifts for Multimodal Learning — arXiv:2610.21403

Source: arXiv (en)
Original: https://arxiv.org/abs/2610.21403
License (as archived): CC0-1.0
Archived: 2026-10-16T10:01:29.651Z
SHA-256 of the archived text: c303d159db192564d310a140b25910bf8deb9e78fef9b13aa1a2452a2d339fae
Stable URL of this page: https://metraipolis.com/documento/dc_578fc183bbcc

How to cite

Beyond Invariance: Factor-wise Selective Treatment of Shifts for Multimodal Learning — arXiv:2610.21403 — arXiv (en), https://arxiv.org/abs/2610.21403 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_578fc183bbcc (Metraipolis archive dc_578fc183bbcc, sha256 c303d159db192564d310a140b25910bf8deb9e78fef9b13aa1a2452a2d339fae)

BibTeX

@misc{dc578fc183bbcc,
  title = {Beyond Invariance: Factor-wise Selective Treatment of Shifts for Multimodal Learning — arXiv:2610.21403},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.21403},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_578fc183bbcc; sha256 c303d159db192564d310a140b25910bf8deb9e78fef9b13aa1a2452a2d339fae},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.21403. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Distribution shifts pose a fundamental challenge to robust multimodal learning. Existing methods typically promote invariance by suppressing environment-dependent variations, implicitly applying a uniform treatment to shifted information. However, shifted information can have different predictive consequences: some contributes little to prediction, some remains reliably informative, and some exhibits environment-dependent predictive effects. This raises a fundamental question: which shifted information should be suppressed, preserved, or attenuated? To address this question, we propose FaShift, a factor-wise shift framework that diagnoses latent representation factors according to their distributional changes, intervention effects, and cross-environment stability. By factorizing multimodal representations and evaluating these complementary properties, FaShift distinguishes non-predictive factors, stably predictive factors, and predictively useful but unstable factors. It then selectively suppresses, preserves, or attenuates their contributions, rather than enforcing uniform invariance. Experiments on MOSI, MOSEI, and CH-SIMS show that FaShift achieves the best ID accuracy and F1 across all three datasets. Compared with the strongest competitors reporting OOD results, it improves OOD accuracy/F1 by 1.5/3.6 points on MOSI and 1.7/2.7 points on MOSEI, while achieving the best reported OOD accuracy and F1 on CH-SIMS.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.21403 — 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 2026-10-16T08:44:47.649Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).