Irreversibility in Non-reciprocal Neural Networks
DIEP seminar by Tuan Pham
In this edition of the DIEP seminar series, Tuan Pham, new DIEP Fellow, will address the question of how entropy production, a measure of time irreversibility, of a complex system with emergent chaotic behaviour is controlled by the heterogeneity in the non-reciprocal interactions among its elements.
Event details from: Irreversibility in Non-reciprocal Neural Networks
- Date
- 26 September 2024
- Time
- 11:00 – 12:00
- Room
- Library
Title
Irreversibility in Non-reciprocal Neural Networks
Abstract
How is entropy production, a measure of time irreversibility, of a complex system with emergent chaotic behaviour controlled by the heterogeneity in the non-reciprocal interactions among its elements? Despite the challenge in computing entropy production within a microscopic theory, the importance of entropy production for the brain’s cognition as recently demonstrated, greatly motivates theoretical work on this question.
In this talk, we address this question using a classic model of random recurrent neural networks that undergoes a dynamical transition from quiescence to chaos at a critical heterogeneity level. By obtaining an exact analytical expression for the averaged entropy production rate, we show how this quantity becomes a constant at the onset of chaos while changing its functional form upon crossing this point. Our work provides the first step to understand not only non-equilibrium phase transitions without broken symmetry but also the connection between dynamics and thermodynamics of complex living systems such as the brain.
If you wish to attend this seminar online, please send an email toc.m.c.demulatier@uva.nl to receive the zoom-link.