Metraipolis › archive › dc_8ddd9d245df3

Beyond receptive fields: sequence-pooled normalization can supply most of a sequence labeler's context — arXiv:2608.18576

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.18576
License (as archived)
CC0-1.0
Archived
2026-10-18T13:53:17.951Z
SHA-256 of the archived text
d6230f3c7a40f2a31d719c795b7350af7be28daec0cb148c47893290de44b5ab
Stable URL of this page
https://metraipolis.com/documento/dc_8ddd9d245df3

How to cite

Beyond receptive fields: sequence-pooled normalization can supply most of a sequence labeler's context — arXiv:2608.18576 — arXiv (en), https://arxiv.org/abs/2608.18576 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_8ddd9d245df3 (Metraipolis archive dc_8ddd9d245df3, sha256 d6230f3c7a40f2a31d719c795b7350af7be28daec0cb148c47893290de44b5ab)
BibTeX
@misc{dc8ddd9d245df3,
  title = {Beyond receptive fields: sequence-pooled normalization can supply most of a sequence labeler's context — arXiv:2608.18576},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.18576},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_8ddd9d245df3; sha256 d6230f3c7a40f2a31d719c795b7350af7be28daec0cb148c47893290de44b5ab},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.18576. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

A convolutional sequence labeler's receptive field is routinely treated as the extent of the model's usable context: it sets dilation schedules, bounds streaming horizons, and underwrites locality claims. However, we show that this can be false: when a normalization layer computes statistics from the current input along the sequence at inference, those statistics open a sequence-spanning path that bypasses the convolutional receptive field to provide global context. We derive this from the layer's Jacobian (the criterion needs no experiment), and what the path carries has a closed form. On a synthetic labeling process with computable optima, the global summary that a sequence-spanning normalization encodes already supplies almost all of what a larger receptive field would buy where labels come in long runs: a network reaching 9 positions comes within 0.009 of the whole-sequence optimum, against a near-chance bound for its reach. Closing the path, by taking the same statistics per position, multiplies what enlarging the receptive field is worth by up to an order of magnitude on simulated genomes at every difficulty level tested and on real 1000 Genomes haplotypes. The same path also confounds attribution: ablating a trained network's receptive-field-enlarging blocks severs part of the path, overstating their contribution 8.3-16.1-fold relative to retraining from scratch. The substitution of normalization for receptive field fades as labels switch more often. Where labels run long, neither the receptive-field justification nor the ablation is wrong about its numbers, but both credit the wrong component.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-20 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG) ; Neural and Evolutionary Computing (cs.NE).
Pagina del lavoro: https://arxiv.org/abs/2608.18576 — 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 2026-10-18T13:50:58.915Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-08-20?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).