The Fast Lane Hypothesis: Von Economo Neurons Implement a Biological Speed-Accuracy Tradeoff

A new computational neuroscience paper by Esila Keskin, titled 'The Fast Lane Hypothesis,' provides the first functional model of Von Economo neurons (VENs). These large, spindle-shaped cells are found only in brain regions linked to complex social behavior in humans, apes, and whales. The hypothesis posits that VENs implement a biological speed-accuracy tradeoff, creating a sparse, fast pathway for rapid social decision-making at the potential cost of deliberate processing accuracy. The model simulates VENs as leaky integrate-and-fire neurons with a 5 ms membrane time constant and sparse connectivity, making them significantly faster than standard pyramidal neurons.

Keskin trained a spiking cortical network of 2,000 neurons on a social discrimination task under three conditions: typical (2% VENs), autism-like (0.4% VENs), and frontotemporal dementia-like (post-training VEN ablation). While all networks achieved the same high final accuracy (99.4%), decision speed varied dramatically. The typical network was significantly faster than the FTD-like model, with median first-spike latencies for VENs occurring 4 ms earlier than for pyramidal cells. This suggests VENs' primary role is modulating decision speed, not learning capacity, offering a mechanistic link to social impairments in autism and FTD where VENs are altered or depleted.