Metraipolis › archive › dc_42e895a96d56

Nyström Attention Matches Full Attention for Cross-Sectional Stock Prediction — arXiv:2609.08106

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.08106
License (as archived)
CC0-1.0
Archived
2026-10-16T13:58:07.457Z
SHA-256 of the archived text
0435a685ea0a5f0e7ec8bd7f5775280d3be911ae152c8d6e11125d57f3929dbf
Stable URL of this page
https://metraipolis.com/documento/dc_42e895a96d56

How to cite

Nyström Attention Matches Full Attention for Cross-Sectional Stock Prediction — arXiv:2609.08106 — arXiv (en), https://arxiv.org/abs/2609.08106 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_42e895a96d56 (Metraipolis archive dc_42e895a96d56, sha256 0435a685ea0a5f0e7ec8bd7f5775280d3be911ae152c8d6e11125d57f3929dbf)
BibTeX
@misc{dc42e895a96d56,
  title = {Nyström Attention Matches Full Attention for Cross-Sectional Stock Prediction — arXiv:2609.08106},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.08106},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_42e895a96d56; sha256 0435a685ea0a5f0e7ec8bd7f5775280d3be911ae152c8d6e11125d57f3929dbf},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.08106. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

MASTER's inter-stock multi-head attention -- the module responsible for modeling cross-sectional stock relationships -- accounts for 42.5% of model parameters and 25% of predictive value. We systematically decompose this module and uncover a surprising structure: the learned attention is near-uniform (perplexity 278/300), yet forcing exact uniformity eliminates all cross-sectional discrimination. Spectral analysis resolves this paradox: the deviation from uniformity is low-rank (effective rank ~65, top-10 modes capture 96.5% of energy), explaining why sparse approximations consistently fail while Nystrom low-rank attention (m=32 landmarks) matches full O(N^2) attention at O(mN) cost -- certified equivalent via TOST at both N=300 (5 seeds, Rank IC p=0.003) and N=800 (10 seeds, Rank IC p=0.034). Additional findings include: (i) attention anti-correlates with return similarity (Spearman rho = -0.614; on the industry-labeled subset, -0.645 unconditionally and -0.627 after controlling for industry, beta, and volatility), suggesting complementarity-seeking rather than correlation mining; (ii) all graph-based alternatives degrade performance, with hard masking worse than complete module removal; and (iii) at N ~ 3,500 with adapted architectures, no cross-stock module (GCN, Nystrom, or MASTER-style pipeline) significantly outperforms a per-stock LSTM baseline (n=4 seeds), indicating that the benefits observed at smaller scales do not trivially transfer. These results establish that the inter-stock attention's value resides in a compressible, dynamic, near-global redistribution that rewards low-rank approximation but resists sparsification.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-09 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG) ; Statistical Finance (q-fin.ST).
Pagina del lavoro: https://arxiv.org/abs/2609.08106 — 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 2026-10-16T13:39:43.020Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-09?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).