Implications:
🧠 Neuroscience — functional rationale for the evolutionary suppression of strong reciprocal motifs.
🤖 NeuroAI — reciprocity as a tunable design parameter in recurrent & neuromorphic networks.

We validated this on empirical connectomes (macaque long-distance, macaque visual cortex, marmoset).

Result: the same pattern.
Strong reciprocity consistently undermines memory and representational richness.

Spectral analysis explains why:

- Higher reciprocity → larger spectral radius (instability risk).
- Narrower spectral gap → less dynamical diversity.
- Lower non-normality → weaker transient amplification.

Together → compressed dynamical range.

Findings (robust across sizes, densities, architectures):
- Increasing reciprocity (link as well as strength reciprocity) reduces memory capacity.
- Representation becomes less diverse (lower kernel rank).
- Effects are strongest in ultra-sparse networks.

The no-strong-loops principle:
Across species (macaque, marmoset, mouse), strong reciprocal (symmetric) connections are rare.

This asymmetry is well known anatomically.
But what are its computational consequences?

Found a #bifurcation (bifurcated ?) #tree outside the Royal Botanic garden, #Sydney 😬

Furthermore, we establish the #scalability of the NRC algorithms by applying them to networks of increasing size with different density and initial reciprocity

We also explored how reciprocity impacts #computation using a memory task in #reservoir_computing frameworks. The influence of graded reciprocity on memory capacity (MC) was found noticeable in both binary and weighted networks

We tested these algorithms on synthetic benchmark networks (random, small-world, modular) and #brain #connectivity maps (#connectomes) across species and reported impact on #spectral properties, #degree_distributions, #community_structure, and #clustering.

Excited to share our Network Reciprocity Control (NRC) algorithms for steering the degree of #asymmetry and #reciprocity in binary and weighted #networks while preserving fundamental network properties.
www.biorxiv.org/content/10.1...
🧵