Activity propagation with Hebbian learning
Will T. Engedal, Róbert Juhász, István A. Kovács · 2025
Read Activity propagation with Hebbian learning by Will T. Engedal, Róbert Juhász, István A. Kovács (2025) 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-2508-21053.txt.
Description
Will Engedal, Róbert Juhász, and István Kovács study, in a 2025 paper, how Hebbian learning changes the spread of activity in the contact process. The contact process is a classic model of spreading, used for infections, brain activity, and population spread. In their version, each successful spread event strengthens or weakens the rate between the two sites involved, in the spirit of learning between brain cells.
The results run against first guesses. Positive reinforcement, which raises activation rates, can destroy the active phase rather than help it. Negative reinforcement, which lowers rates, can turn an inactive system globally active. In two dimensions and above, weakening links can also leave patches that never activate, producing two critical points in place of one. Strong reinforcement can also build loops of activity that trap the spread, like ants marching in a circle.
Catalog record
| CREATOR | Will T. Engedal, Róbert Juhász, István A. Kovács |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 61 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-21053