Linear-Readout Floors and Threshold Recovery in Computation in Superposition

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Borobia, Hector, Seguí-Mas, Elies, Tormo-Carbó, Guillermina
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866913083287404544
author Borobia, Hector
Seguí-Mas, Elies
Tormo-Carbó, Guillermina
author_facet Borobia, Hector
Seguí-Mas, Elies
Tormo-Carbó, Guillermina
contents Two recent approaches to computation in superposition reach different recursive capacity regimes: Hänni et al. certify $\tilde{O}(d^{3/2})$ computable features in width $d$ via an approximate-linear recursive template, while Adler and Shavit reach near-quadratic capacity (up to logarithmic factors) using thresholded Boolean recovery. The main contribution of this paper is conceptual: we argue these results are not contradictory because they maintain different interface invariants, and we formalize the distinction. As a tool, we record a rank-trace Welch-type lower bound for biorthogonal linear readouts: for $F \gg d$, the worst-case off-diagonal cross-talk of any unit-diagonal linear readout is $Ω(d^{-1/2})$, and the bound is tight on average for unit-norm tight frames. At quadratic feature load $F=d^2$, random-support threshold recovery succeeds for sparsities $s=O(d/\log d)$, while linear readouts still incur $Ω(s/d)$ average per-coordinate squared error on Bernoulli sparse states. Matching the Welch floor against the published tolerance of the Hänni correction layer explains the $d^{3/2}$ scale as a compatibility threshold for that template, not a universal upper bound. Robust nonlinear reset beyond the Hänni template is left open.
format Preprint
id arxiv_https___arxiv_org_abs_2605_01192
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Linear-Readout Floors and Threshold Recovery in Computation in Superposition
Borobia, Hector
Seguí-Mas, Elies
Tormo-Carbó, Guillermina
Machine Learning
Information Theory
I.2.6; F.1.1
Two recent approaches to computation in superposition reach different recursive capacity regimes: Hänni et al. certify $\tilde{O}(d^{3/2})$ computable features in width $d$ via an approximate-linear recursive template, while Adler and Shavit reach near-quadratic capacity (up to logarithmic factors) using thresholded Boolean recovery. The main contribution of this paper is conceptual: we argue these results are not contradictory because they maintain different interface invariants, and we formalize the distinction. As a tool, we record a rank-trace Welch-type lower bound for biorthogonal linear readouts: for $F \gg d$, the worst-case off-diagonal cross-talk of any unit-diagonal linear readout is $Ω(d^{-1/2})$, and the bound is tight on average for unit-norm tight frames. At quadratic feature load $F=d^2$, random-support threshold recovery succeeds for sparsities $s=O(d/\log d)$, while linear readouts still incur $Ω(s/d)$ average per-coordinate squared error on Bernoulli sparse states. Matching the Welch floor against the published tolerance of the Hänni correction layer explains the $d^{3/2}$ scale as a compatibility threshold for that template, not a universal upper bound. Robust nonlinear reset beyond the Hänni template is left open.
title Linear-Readout Floors and Threshold Recovery in Computation in Superposition
topic Machine Learning
Information Theory
I.2.6; F.1.1
url https://arxiv.org/abs/2605.01192