主讲人：Antonio Politi（意大利国家研究委员会）

题目：Collective phenomena in neural networks

时间：2023年10月11日下午14:30

地点：理工楼1201

报告摘要：

The behavior of oscillator networks can be highly complex, leading to the manifestation of a wealth of nonintuitive collective properties. First, I discuss the emergence of macroscopic irregular (but deterministic) dynamics, showing that it may appear even in relatively simple models consisting of globally coupled homogeneous networks of phase oscillators. Altogether, the resulting regime can be viewed as an instance of partial synchronization. Next, I focus on the so-called balanced state, where the neurons fire irregularly with a finite spiking rate in spite of strong (virtually infinite) synaptic currents. This regime is considered to be very relevant within computational neuroscience, since it is believed to describe the resting state of the mammalian brain.The typical setups found in the literature require either a non-generic setting of the coupling parameters, or the unjustified presence of strong external currents. I discuss a slightly different setup, which is able to overcome such limitations. It consists of pulse-coupled oscillators (such as the leaky integrate and fire neurons), where the nonlinearity induced by the resource finiteness provides a robust self-adjusting mechanism, which ensures the stability of a balanced state in generic conditions.

个人简介：

Antonio Politi, PhD in Physics with a Thesis on Optical Bistability.International Workshops and Conferences, more than 70 invitations; Seminars and colloquia, More than 40 invitations. Have over 195 papers published in peer-reviewed journals (17 of his publications have been cited more than 100 times).