Geometric Attention A Regime-Explicit Operator Semantics for Transformer Attention
Luis Rosario Freytes · 2026
Read Geometric Attention A Regime-Explicit Operator Semantics for Transformer Attention by Luis Rosario Freytes (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-11618.txt.
Introduction
Luis Rosario Freytes defines attention in geometric, regime-explicit terms. The 2026 preprint names the frame Geometric Attention, or GA. Transformer attention is usually given as one fixed formula. GA instead specifies a layer by four independent inputs.
The first input is a finite carrier of addressable indices. The second is an evidence-kernel rule for weights. Masked proto-scores and a link induce nonnegative weights there. The third input is a probe family of admissible observables. The fourth is an anchor and update rule for kernels. It selects which representative kernel applies and how. The split makes hidden choices in attention explicit. Designers can then vary one choice at a time.
Catalog record
| CREATOR | Luis Rosario Freytes |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 173 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2601-11618