Metraipolis › archive › dc_5fa620341776
Relational Composition as Parallel Transport: What Per-Entity Frames Force in Attention — arXiv:2610.34945
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.34945
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-27T08:02:24.789Z
- SHA-256 of the archived text
03ef6424086ac745da2e6044b7b7bf663d20cba403fce9a9dc1904385c423fd9- Stable URL of this page
https://metraipolis.com/documento/dc_5fa620341776
How to cite
Relational Composition as Parallel Transport: What Per-Entity Frames Force in Attention — arXiv:2610.34945 — arXiv (en), https://arxiv.org/abs/2610.34945 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_5fa620341776 (Metraipolis archive dc_5fa620341776, sha256 03ef6424086ac745da2e6044b7b7bf663d20cba403fce9a9dc1904385c423fd9)
BibTeX
@misc{dc5fa620341776,
title = {Relational Composition as Parallel Transport: What Per-Entity Frames Force in Attention — arXiv:2610.34945},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.34945},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_5fa620341776; sha256 03ef6424086ac745da2e6044b7b7bf663d20cba403fce9a9dc1904385c423fd9},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.34945. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
A dot product means something only when both vectors share a coordinate system. Content words have one; a fresh entity does not: rotating its coordinates alone changes nothing. We take this per-entity frame freedom as a symmetry and ask what it forces in attention. For orthogonal transports, invariant scores depend only on norms and the transported inner product $\langle x_i,T_{ij}x_j\rangle$; equivariant messages lie in the span of $x_i$ and $T_{ij}x_j$. Endpoint channel count bounds transport rank, so with one vector per entity relations must be supplied; standard attention is the flat sector of a shared frame. Representation theory attaches an integer to each mechanism; the readout decides which. With orthogonal transitions, a convex-cell readout composes a finite group exactly only at or above its smallest faithful real representation dimension. At this minimum, exactness forces a faithful representation. An even readout can decode a cover at smaller width, and the gap is unbounded. Below its threshold, the part of the group a mechanism cannot see sets an accuracy ceiling computable before training. 17/17 fitting seeds at minimal width passed tests through length 1000, versus 7/18 at twice the width; 12 models carry correctness certificates to explicit lengths, from the weights alone. On $C_3\times A_4$ (faithful width 5), descent at width 4 learned the 72-state double cover under sign-blind attention (6/48 seeds; monotone 0/48); on a frozen cover the cross-entropy-optimal affine head is uniform, a quadratic head exact. On parsed CLUTRR, transported composition beats a matched transformer by 41 points at 4-10 hops, but forcing isometries costs 65 points; on StepGame's commutative algebra, isometries help. Static embeddings gave each word one vector; contextual models gave each occurrence its own; per-entity frames make each relation an operator, the unit in which relations compose.
Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI); Formal Languages and Automata Theory (cs.FL).
Pagina del lavoro: https://arxiv.org/abs/2610.34945 — 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 2026-10-27T07:55:30.952Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).