Metraipolis › archive › dc_dbe989b3dcbf

On the Limits of Scale: Inverse Scaling in Vision-Language Models — arXiv:2610.27360

Source: arXiv (en)
Original: https://arxiv.org/abs/2610.27360
License (as archived): CC0-1.0
Archived: 2026-10-21T07:07:41.863Z
SHA-256 of the archived text: 36a8dfc70f7690d6e9bf6020439053e153b77d2c825d0e80469685b9a9669c2b
Stable URL of this page: https://metraipolis.com/documento/dc_dbe989b3dcbf

How to cite

On the Limits of Scale: Inverse Scaling in Vision-Language Models — arXiv:2610.27360 — arXiv (en), https://arxiv.org/abs/2610.27360 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_dbe989b3dcbf (Metraipolis archive dc_dbe989b3dcbf, sha256 36a8dfc70f7690d6e9bf6020439053e153b77d2c825d0e80469685b9a9669c2b)

BibTeX

@misc{dcdbe989b3dcbf,
  title = {On the Limits of Scale: Inverse Scaling in Vision-Language Models — arXiv:2610.27360},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.27360},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_dbe989b3dcbf; sha256 36a8dfc70f7690d6e9bf6020439053e153b77d2c825d0e80469685b9a9669c2b},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.27360. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Scaling laws suggest that increasing model size, training data, and compute generally yields monotonic performance improvements. While this often holds at the aggregate level, we present systematic evidence that vision-language models (VLMs) exhibit inverse scaling at the subtask level: specific question categories on which larger models perform worse than smaller counterparts within the same model family. We evaluate 19 models from four distinct model families across eight multimodal benchmarks (nine evaluation splits), answering 29,386 questions per model, 558,334 model-question evaluations in total. Using per-question pairwise analysis across all model pairs in each family, we identify 375 subtask-model-pair combinations exhibiting inverse scaling, spanning 4,342 unique queries where a smaller model answers correctly but a larger model fails, and we find that these patterns replicate across independently developed architectures. We identify 13 subtask-dataset pairs that show inverse scaling across multiple families, consistent with the phenomenon being task-driven rather than model-specific. The most affected domains in our study include multi-image spatial reasoning, visual perception, and university-level STEM knowledge, with accuracy gaps exceeding 20 percentage points in several cases, while diagram understanding appears comparatively resistant. These findings challenge the assumption that increasing model size uniformly improves multimodal capabilities and suggest that model selection should be task-aware.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.27360 — 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 2026-10-21T07:05:43.553Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).