Abstract. Artificial neural networks achieve remarkable performance through learned weight distributions over largely arbitrary topologies, yet biological neural circuits evolved under strict metabolic and behavioral constraints over millions of years. Whether this evolved wiring encodes computational priors that confer advantages in learning efficiency, generalization, or representational sparsity over randomly initialized architectures remains an open question. We present a controlled comparison between a Spiking Neural Network whose recurrent synaptic topology is constrained to the Drosophila melanogaster olfactory circuit as mapped by the FlyWire connectome (Shiu et al., 2024) and three baselines: a degree-matched randomly-wired SNN, a sparsity-matched MLP, and a fully-connected parameter-matched MLP. All models are trained on the DoOR olfactory receptor response dataset for multi-class odor identity classification using surrogate gradient descent. We hypothesize that biologically-derived synaptic topology encodes an inductive bias toward more efficient or stable olfactory representations than randomly-initialized architectures of equivalent capacity. We evaluate convergence rate, peak classification accuracy, spike sparsity, and training stability across five-fold cross-validation and multiple random seeds. All code, connectome extraction pipelines, and trained weights are released to support reproducibility and extension to other FlyWire circuit subgraphs.

This work reflects MIRE's commitment to rigorous, open, and honest science — including the publication of clear negative results, which are as valuable to the field as positive ones. By releasing our full code, connectome-extraction pipeline, and trained weights, we aim to support reproducibility and invite others to extend the work to new circuits and learning rules.

Read the full paper below and explore the code on GitHub — https://github.com/VickM12/flywire-olfactory-snn