Fine-grained system identification of nonlinear neural circuits

Data used in Bagherian et al 2021, published at KDD. Code utilizing this dataset can be found at github.com/dbagherian/neural-sys-id We study the problem of sparse nonlinear model recovery of high dimensional compositional functions. Our study is motivated by emerging opportunities in neuroscience to recover fine-grained models of biological neural circuits using collected measurement data.

Guided by available domain knowledge in neuroscience, we explore conditions under which one can recover the underlying biological circuit that generated the training data. Our results suggest insights of both theoretical and practical interests. Most notably, we find that a sign constraint on each layer of weights is a sufficient condition for system recovery (in addition to other more standard sufficient conditions, which we establish both theoretically with an identifiability guarantee and empirically on simulated biological circuits.