Holographic functions and neural networks
Balázs Szegedy · 2026
Read Holographic functions and neural networks by Balázs Szegedy (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-2605-22666.txt.
Description
Szegedy compares three ways of calling a Boolean function simple. The first is sampling recovery from few coordinates. The second is closeness to a low degree polynomial in few linear forms. The third is nearness to a neural net output. He names the sampling trait holography and proves the three views equivalent up to shifts in parameters. The step from sampling to the second view leans on a form of hypergraph regularity. The result ties a sampling intuition, a shape account and a computational model of complexity into one equivalence.
Catalog record
| CREATOR | Balázs Szegedy |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 44873 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.22666