Purely STDP-based assembly dynamics: Stability, learning, overlaps, drift and aging

Abstract

Memories may be encoded in the brain via strongly interconnected groups of neurons, called assemblies. The concept of Hebbian plasticity suggests that these assemblies are generated through synaptic plasticity, strengthening the recurrent connections within select groups of neurons that receive correlated stimulation. To remain stable in absence of such stimulation, the assemblies need to be self-reinforcing under the plasticity rule. Previous models of such assembly maintenance require additional mechanisms of fast homeostatic plasticity often with biologically implausible timescales. Here we provide a model of neuronal assembly generation and maintenance purely based on spike-timing-dependent plasticity (STDP) between excitatory neurons. It uses irregularly and stochastically spiking neurons and STDP that depresses connections of uncorrelated neurons. We find that assemblies do not grow beyond a certain size, because temporally imprecisely correlated spikes dominate the plasticity in large assemblies. Assemblies in the model can be learned or spontaneously emerge. The model allows for prominent, stable overlap structures between static assemblies. Further, assemblies can drift, particularly according to a novel, transient overlap-based mechanism. Finally the model indicates that assemblies grow in the aging brain, where connectivity decreases.

Co-existence of synaptic plasticity and metastable dynamics in a spiking model of cortical circuits

“…We also note that our results on remapping of sensory stimuli is different from the reorganization of clusters described in (Manz and Memmesheimer, 2023), which has been related to representational drift (Ziv et al, 2013;DeNardo et al, 2019;Rule et al, 2019;Masset et al, 2022). In our model, new clusters form due to stimulation by a new set of stimuli rather than drifting of single neuron representations due to the noisy dynamics of plasticity as it occurs in (Manz and Memmesheimer, 2023).…”

“…A similar result was obtained with the plasticity rule of (Litwin-Kumar and Doiron, 2014), which however requires inhibitory plasticity and synaptic renormalization. More recently, (Manz and Memmesheimer, 2023) have proposed a pure STDP plasticity rule that, similarly to our rule, does not require inhibitory plasticity or synaptic renormalization. However, (Manz and Memmesheimer, 2023) focus on the generation of stable clusters producing persistent activity rather than metastable dynamics.…”

“…More recently, (Manz and Memmesheimer, 2023) have proposed a pure STDP plasticity rule that, similarly to our rule, does not require inhibitory plasticity or synaptic renormalization. However, (Manz and Memmesheimer, 2023) focus on the generation of stable clusters producing persistent activity rather than metastable dynamics. Their persistent activity has the signature of fast Poisson noise due to the stochastic nature of their model neurons, but lacks the slow fluctuations characteristic of ongoing metastable dynamics (Litwin-Kumar and Doiron, 2012).…”

Synaptic motility and functional stability in the whisker cortex

“…This noise will cause the synaptic weights to fluctuate around a point attractor or induce a random walk on a manifold attractor. Consequently, it has been argued that activity-independent plasticity can generate 'drifting representations' [58][59][60][61][62][63][64]. That is, the neural representation of an external stimulus will not be fixed in time.…”

Drift of neural ensembles driven by slow fluctuations of intrinsic excitability

“…However, the mechanisms underlying the emergence of drift and its relevance for the neural computation are not known. Drift is often thought to arise from variability of internal states ( Sadeh and Clopath, 2022 ), neurogenesis ( Rechavi et al, 2022 ; Driscoll et al, 2017 ) or synaptic turnover ( Attardo et al, 2015 ) combined with noise ( Kossio et al, 2021 ; Manz and Memmesheimer, 2023 ). On the other hand, excitability might also play a role in memory allocation ( Zhou et al, 2009 ; Mau et al, 2020 ; Rogerson et al, 2014 ; Silva et al, 2009 ), so that neurons having high excitability are preferentially allocated to memory ensembles ( Cai et al, 2016 ; Rashid et al, 2016 ; Silva et al, 2009 ).…”