Memory Uncertainty Relation and Harmonic Memory in Random Recurrent Networks
Taichi Haruna, Kohei Nakajima · 2026
Read Memory Uncertainty Relation and Harmonic Memory in Random Recurrent Networks by Taichi Haruna, Kohei Nakajima (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-2605-24628.txt.
Description
Taichi Haruna and Kohei Nakajima prove a lower bound on the short-term memory of a dynamical system. It trades memory capacity against the size of input-driven state swings. A readout weight can attain the bound, yielding weaker harmonic memory. They test the inequality across reservoir systems with input noise. Equality holds exactly sometimes, only in the limit sometimes, and never otherwise. Under regularization, a certain noise strength adds memory, explained as noise-induced memory.
Catalog record
| CREATOR | Taichi Haruna, Kohei Nakajima |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 59109 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.24628