Abstract

Please join us for our fortnightly journal club online via zoom where two presenters will jointly present a topic together. The next topic is ‘optimal spike coding’ presented by David Liu and Jeroen Olieslagers.

Zoom information: https://us02web.zoom.us/j/84958321096?pwd=dFpsYnpJYWVNeHlJbEFKbW1OTzFiQT09 Meeting ID: 849 5832 1096 Passcode: 506576

In this journal club, we will explore the topic of optimal spike coding. This line of research provides an alternative view on neural coding. In particular, individual spikes carry significant information, and stochasticity arising in networks is not merely noise as viewed from a conventional Poisson rate model perspective. Different versions of such spiking dynamics have been explored in the literature, but the main idea underpinning this framework is simple: each neuron only fires whenever the error in decoding the encoded signal exceeds some threshold. This assigns functional meaning to spiking thresholds, membrane potentials and synaptic connections. When applied to recurrently connected networks, activity exhibits Poisson-like statistics and mirrors the asynchronous firing states reported for spontaneous activity in the cortex. We will go through the following papers that introduce this framework, as well as extensions to probabilistic representations of uncertainty:

David: 1. Bourdoukan, Ralph, et al. “Learning optimal spike-based representations.” Advances in neural information processing systems 25 (2012): 2285-2293.

2. Boerlin, Martin, and Sophie Denève. “Spike-based population coding and working memory.” PLoS Comput Biol 7.2 (2011): e1001080.

Jeroen: 3. Savin, Cristina, and Sophie Deneve. “Spatio-temporal Representations of Uncertainty in Spiking Neural Networks.” NIPS . Vol. 27. 2014.

4. Brendel W, Bourdoukan R, Vertechi P, Machens CK, Denéve S. “Learning to represent signals spike by spike.” PLoS computational biology. 2020 Mar 16;16(3):e1007692.