Representational drift as the consequence of ongoing memory storage

Abstract

Memory systems with biologically constrained synapses have been the topic of intense theoretical study for over thirty years. Perhaps the most fundamental and far-reaching finding from this work is that the storage of new memories implies the partial erasure of already-stored ones. This overwriting leads to a decorrelation of sensory-driven activity patterns over time, even if the input patterns remain similar. Representational drift (RD) should therefore be an expected and inevitable consequence of ongoing memory storage. We tested this hypothesis by fitting a network model to data from long-term chronic calcium imaging experiments in mouse hippocampus. Synaptic turnover in the model inputs, consistent with the ongoing encoding of new activity patterns, accounted for the observed statistics of RD. This mechanism also provides a parsimonious explanation for the recent finding that RD in CA1 place cells has two distinct components: one which depends only on the passage of time, and another which depends on the time spent exploring a given environment. Furthermore, in the context of ongoing learning, the drift rate of any one memory depends on its repetition rate, a mechanism which can reproduce the diverse effects of experience on drift found in experiment. Our results suggest that RD should be observed wherever neuronal circuits are involved in a process of ongoing learning or memory storage.

Single field evolution rule governs the dynamics of representational drift in mouse hippocampal dorsal CA1 region

“…SFER also explains how the average lifespan of place fields gradually lengthens, a key feature for representational stabilization (Figures 5J-W). Its convergent dynamics resonate with recent computational frameworks proposing cross-day RD as a consequence of continuous representation refinement 56,57 , a distinct dynamics compared to RD observed in other brain regions 31,58 , suggesting that SFER is confined to the entorhinal-hippocampal circuit and results from characteristics unique to these circuits. However, it is essential to emphasize several considerations regarding SFER: 1) The validation of SFER is currently limited to a two-month period, and its applicability beyond this timeframe remains uncertain.…”