Metraipolis › archive › dc_ebe5963e1c92

MJev-Spatial: System One Spatial Intelligence via Native Attention Matching — arXiv:2610.22276

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.22276
License (as archived)
CC0-1.0
Archived
2026-10-16T10:39:09.104Z
SHA-256 of the archived text
c4404c99b52fd611175bee8fb7d70eb80d16b15556eb3c8983085a184033b933
Stable URL of this page
https://metraipolis.com/documento/dc_ebe5963e1c92

How to cite

MJev-Spatial: System One Spatial Intelligence via Native Attention Matching — arXiv:2610.22276 — arXiv (en), https://arxiv.org/abs/2610.22276 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ebe5963e1c92 (Metraipolis archive dc_ebe5963e1c92, sha256 c4404c99b52fd611175bee8fb7d70eb80d16b15556eb3c8983085a184033b933)
BibTeX
@misc{dcebe5963e1c92,
  title = {MJev-Spatial: System One Spatial Intelligence via Native Attention Matching — arXiv:2610.22276},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.22276},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ebe5963e1c92; sha256 c4404c99b52fd611175bee8fb7d70eb80d16b15556eb3c8983085a184033b933},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.22276. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Spatial intelligence lets an agent perceive and act in the physical world, and many of its judgments must be made at interaction rates: which spatial relation holds, how far an object is, or where to place it. Vision-language models (VLMs) produce these answers by free-form text generation, paying for autoregressive decoding and answer parsing even though the output space is compact: one of a few options, a scalar, or a 2D point. In light of this, we introduce MJev-Spatial, a System 1 multimodal decision model that tackles all three types of spatial understanding tasks within a unified classification framework. Specifically, MJev-Spatial treats each candidate as a class and scores it using the backbone's pretrained query and key projections, without adding parameters. The candidates are the given options for choice tasks, or value bins and image regions refined through coarse-to-fine classification for numerical and point prediction, respectively. At inference, each classification decision takes a single forward pass and generates no answer tokens. To train on heterogeneous spatial data efficiently, we supervise all coarse-to-fine stages jointly with teacher forcing in one forward pass, and combine this with length-aware global batching and cross-GPU load balancing, raising training throughput by 30.8% over the original batching strategy. Built on Qwen3.5-4B and trained on 12.3M question-answer pairs converted from existing spatial datasets, MJev-Spatial achieves the highest overall score (65.64%) among eight tested models, ranks first on seven of twelve benchmarks, and outperforms its unmodified autoregressive base model on all twelve, while running 2.5 times faster than the fastest autoregressive baseline and 5.6 times faster than the base model.

Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.CV, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2610.22276 — 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 2026-10-16T08:52:02.732Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).