Bayesian reconstruction of memories stored in neural networks from their connectivity
This work addresses a key challenge in connectomics for neuroscientists, but it is incremental as it builds on existing attractor network models.
The authors tackled the problem of reconstructing stored information from synaptic connectivity in recurrent neural networks, developing a Bayesian inference algorithm that performs comparably to PCA across three models and analyzing its theoretical limits.
The advent of comprehensive synaptic wiring diagrams of large neural circuits has created the field of connectomics and given rise to a number of open research questions. One such question is whether it is possible to reconstruct the information stored in a recurrent network of neurons, given its synaptic connectivity matrix. Here, we address this question by determining when solving such an inference problem is theoretically possible in specific attractor network models and by providing a practical algorithm to do so. The algorithm builds on ideas from statistical physics to perform approximate Bayesian inference and is amenable to exact analysis. We study its performance on three different models, compare the algorithm to standard algorithms such as PCA, and explore the limitations of reconstructing stored patterns from synaptic connectivity.