von Economo neurons (VENs) are large bipolar projection neurons found exclusively in the anterior cingulate cortex (ACC) and frontal insula of species with complex social cognition, including humans, great apes, cetaceans, and elephants. Their selective depletion in frontotemporal dementia (FTD) and altered development in autism implicate them in rapid social decision-making, yet no computational model of VEN function has previously existed. We introduce the Fast Lane Hypothesis: VENs implement a biological speed-accuracy tradeoff (SAT) by providing a sparse, fast projection pathway that enables rapid social decisions at the cost of deliberate processing accuracy. We model VENs as fast leaky integrate-and-fire (LIF) neurons (tau=5ms, fan-in=8) alongside standard pyramidal neurons (tau=20ms, fan-in=80) within a spiking cortical circuit (N_pyr=2,000) trained on a social discrimination task. Networks are evaluated under three clinically motivated conditions across 10 independent random seeds: typical (2% VENs), autism-like (0.4% VENs), and FTD-like (post-training VEN ablation). All configurations achieve equivalent asymptotic classification accuracy (~99.4%), consistent with the prediction that VENs modulate decision speed rather than representational capacity. VENs produce median first-spike latencies 4ms earlier than pyramidal neurons, providing direct evidence of a fast signalling pathway. At a fixed decision threshold of theta=3, the typical condition is significantly faster than FTD-like (t=-23.31, p<0.0001), while the autism-like condition is intermediate (mean RT=26.91+/-9.01ms vs. typical 20.70+/-2.02ms; p=0.078, trending). To our knowledge, this is the first computational model that asks what a von Economo neuron actually computes.
Figures & tables
fven (%)
Nven
Val. acc. (%)
Plateau (ep.)
0.0
0
99.20±0.00
2.0
0.5
10
99.40±0.16
1.0
1.0
20
99.47±0.09
1.3
2.0
40
99.47±0.09
1.0
3.0
60
99.40±0.16
1.3
5.0
100
99.33±0.09
1.0
Table 1: VEN fraction sweep ( Npyr=2,000 , 3 seeds, mean ± SD). All configurations converge to ∼ 99.4% accuracy, confirming VENs modulate speed rather than capacity.
Figure 1: VEN fraction sweep ( Npyr=2,000 , 3 seeds, mean ± SD). Validation accuracy is flat across all VEN fractions from 0–10%, confirming that VENs modulate decision speed rather than representational capacity. The dashed line marks the biological optimum of ∼ 2%.
Figure 2: Clinical conditions at fixed θ=3 ( Npyr=2,000 , 10 seeds, mean ± SD). Left: Mean reaction time. Typical is fastest ( 20.70±2.02 ms), FTD-like is slowest ( 32.40±0.52 ms; p<0.0001 ), autism-like is intermediate ( 26.91±9.01 ms). The large SD for autism reflects seed-to-seed variability in pyramidal compensation. Right: Decision accuracy is near-ceiling ( > 99%) across all conditions, confirming VENs modulate speed, not capacity.
Condition
fven
Mean RT (ms)
Acc. (%)
Typical
2.0%
20.70±2.02
99.72±0.15
Autism-like
0.4%
26.91±9.01
99.70±0.23
FTD-like
2.0% †
32.40±0.52
99.27±0.11
Paired t -tests (RT):
Typical vs. FTD: t=−23.31 , p<0.0001 [***]
Typical vs. Autism: t=−1.99 , p=0.078 [ns, trending]
Table 2: Clinical conditions at θ=3 ( Npyr=2,000 , 10 seeds, mean ± SD). Statistical comparisons via paired t -test.
Figure 3: Per-seed reaction time distribution for autism-like condition ( Npyr=2,000 , fven=0.4% , 10 seeds, fixed θ=3 ). Histogram reveals bimodal distribution with two clusters: successful compensation (6 seeds, mean RT ≈ 20.3 ms) and incomplete compensation (4 seeds, mean RT ≈ 36.8 ms). This seed-to-seed variability reflects the stochasticity of pyramidal pathway reorganisation during training with reduced VEN input, consistent with heterogeneous autism phenotypes.
Figure 4: Clinical conditions under adaptive threshold ( Δ=1 ) ( Npyr=2,000 , 10 seeds, mean ± SD). Left: Mean RT is compressed relative to fixed θ=3 , but the ordering is preserved: typical fastest, FTD-like slowest. Right: Accuracy under the adaptive protocol is lower for faster conditions, directly demonstrating the speed-accuracy tradeoff. FTD-like achieves the highest accuracy precisely because it decides more slowly.
fven (%)
2 cls
4 cls
6 cls
8 cls
12 cls
16 cls
Chance
50.0
25.0
16.7
12.5
8.3
6.3
0.1
50.0
25.0
16.7
12.5
9.8
7.5
0.5
50.0
25.0
16.7
12.5
10.2
7.2
1.0
50.0
25.0
16.7
12.5
10.5
7.1
2.0
50.0
25.0
16.7
12.5
10.0
7.4
5.0
50.0
25.0
16.7
12.5
9.8
6.9
Table 3: Evolution experiment: validation accuracy (%) by social complexity and VEN fraction (single seed, 15 epochs). Networks achieve at or near chance for Nclasses≤8 , indicating insufficient training.
Figure 5: Preliminary evolutionary analysis. Model-optimal VEN fraction vs. social task complexity, overlaid with empirical VEN densities for macaque (0%), gorilla (0.5%), chimpanzee (0.9%), and human (2.0%) ( Butti et al., 2013 ; Allman et al., 2010 ) . The qualitative correspondence — higher task complexity requires more VENs — motivates future experiments with sufficient training. Presented as exploratory only.
Von Economo neurons (VENs) are selectively lost in behavioural-variant frontotemporal dementia (bvFTD) and reduced in autism spectrum conditions (ASC), yet their computational role in social learning remains unexplained. We train a spiking neural network (the VENCircuit) embedding VEN-like projection neurons (K=40, 2% of total) in a recurrent pyramidal circuit across 50 matched random initialisations with and without VENs. The network is trained on a controlled binary classification task; we make no claim to model social cognition directly. VEN-intact networks converged in 49/50 cases (98%) versus 35/50 (70%) for VEN-ablated networks (Fisher's exact OR=21.0, 95% CI 2.7-167, p=8.7e-5). Failed ablated networks showed complete absence of learning, inconsistent with a speed-of-learning account. Phase-ablation experiments show VEN removal is most disruptive during mid-training (epochs 5-25), when a co-adaptive dependency forms in the pyramidal circuit. We derive a formal account showing VENs provide a direct gradient pathway immune to Jacobian instabilities affecting the recurrent circuit. Inference-time VEN ablation caused a significant performance drop (Wilcoxon p=0.022), ranging from no change (16/20 networks) to catastrophic collapse (0.989 to 0.620). VENs function as acquisition scaffolds whose developmental absence produces stochastic learning failure - a computational analogue of variable social skill acquisition in ASC - with falsifiable predictions for organoid and electrophysiology studies.
Esila Keskin
School of Computing and Creative Technologies, University of the West of England, Bristol, UK
Cortical neurons fire sparsely -- often fewer than one spike per sensory window -- so a neuron's spike count in that window carries little information, and its spike timing can carry more. That conduction delays convert firing order into synchrony is long established. What governs which class of temporal feature a neuron detects -- one volley of coincident input, or two in a particular order -- has not been examined. We propose a delay-signature framework in which the axonal conduction delays converging on a dendritic branch or a small neuron's soma constitute a physical key: only input sequences whose spike-time differences the delays compensate arrive synchronously, and coincidence detection, a threshold on input summed within an inhibition-compressed window, converts that synchrony into an all-or-none output. First, in simulations of an integrator-neuron model, a single physical scalar -- the dispersion of the delay set -- moves a population from event to order-selective sequence detection. The transition is emergent under random delays and connectivity, and an exact calculation reproduces it: at narrow dispersion, sequence detectors do not exist, and the dispersion at which they overtake event detectors tracks the inter-event interval. This maps a computational distinction onto the anatomical one between myelinated and unmyelinated projections, making myelination a switch on what a neuron computes. Second, the same dispersion sets the code's limits: it bounds the longest codable interval and fixes an absolute timing tolerance of order a millisecond, the same at every dispersion. Third, that millisecond window and the conduction velocity of afferents inside cortex together set column diameter, and column anatomy predicts the window measured in mouse barrel cortex. One anatomically measurable parameter thus shapes what a neuron detects and sets the limits of what it can represent.
Cheng Bi, Jipeng Sun
Independent Researcher · Department of Computer Science, Princeton University
Exploiting temporal information in datasets with high temporal dependencies (such as spoken keyword recognition) by spiking models is still an open problem. Promising frameworks include Time-To-First coding (TTFS) and the inclusion of synaptic delays into learning. In TTFS, neurons are forced to spike at most once, forcing the network to found its learning on the spiking times of neurons rather than their spiking rate. Some models use the TTFS framework to directly compute the spiking time of Integrate-and-Fire (IF) or Leaky Integrate-and-Fire (LIF) neurons. However, most models built in the TTFS framework have not been tested against datasets containing temporal information. Little is known about the latency of these models on temporal datasets, where the latency is the time taken by the network to correctly classify a sample. In this work, we study the behaviour of Delay Neural Networks (DeNN) models, i.e. models that directly compute the spiking time of neurons by bypassing the membrane potential and thresholding mechanisms of spiking neurons. We show how to analytically control their latency, and empirically study the latency-accuracy tradeoff in a dynamic temporal dataset.