Metraipolis › archive › dc_8100b62371ec

Rotation Is Token-Blind: A Structural Criterion for Post-Training Quantization of Feed-Forward Geometry Transformers — arXiv:2611.07955

Source: arXiv (en)
Original: https://arxiv.org/abs/2611.07955
License (as archived): CC0-1.0
Archived: 2026-11-06T14:53:37.909Z
SHA-256 of the archived text: 5fdcf224db6049ba6a441880367077d2f6a3745de9a384703e42d30e5bb0cfd5
Stable URL of this page: https://metraipolis.com/documento/dc_8100b62371ec

How to cite

Rotation Is Token-Blind: A Structural Criterion for Post-Training Quantization of Feed-Forward Geometry Transformers — arXiv:2611.07955 — arXiv (en), https://arxiv.org/abs/2611.07955 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_8100b62371ec (Metraipolis archive dc_8100b62371ec, sha256 5fdcf224db6049ba6a441880367077d2f6a3745de9a384703e42d30e5bb0cfd5)

BibTeX

@misc{dc8100b62371ec,
  title = {Rotation Is Token-Blind: A Structural Criterion for Post-Training Quantization of Feed-Forward Geometry Transformers — arXiv:2611.07955},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.07955},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_8100b62371ec; sha256 5fdcf224db6049ba6a441880367077d2f6a3745de9a384703e42d30e5bb0cfd5},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.07955. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Feed-forward 3D models use Transformers to predict camera poses, depth, and structure in one pass, avoiding iterative multi-view optimization. Long sequences and deep attention incur high compute and memory costs, motivating Post-Training Quantization (PTQ) for deployment. Existing PTQ methods mainly suppress channel-wise outliers by smoothing or orthogonal rotation. Geometry Transformers introduce a different challenge: heterogeneous token types share one activation quantization grid. A few special tokens may exert high leverage over global outputs and be updated by only a subset of blocks. Token-blind transforms, which apply the same transformation to every token, reduce overall quantization error but cannot remove differentiated degradation across token types. We propose QMapAnything, a structure-aware PTQ framework. We cast floating-point-preserving rotation and smoothing as a token-blind reparameterization group, and prove that they yield common-mode gains while leaving relative token SNR unchanged. We introduce a statistic $\Gamma$ that combines a special token's quantization disadvantage and output leverage to predict whether differentiated bias correction is needed beyond rotation. Experiments on geometric benchmarks show that W4A8 closely matches full-precision performance. Under W8A8, peak memory falls by $2.2$--$3.1\times$ versus FP32, with up to $3.1\times$ speedup on real hardware. These results provide a principled and efficient approach to deploying feed-forward 3D reconstruction models in resource-constrained edges.

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/2611.07955 — 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 2026-11-06T14:25:42.622Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).