neural-receptive-fields-scale-free-geometry

name	neural-receptive-fields-scale-free-geometry
description	Geometric framework for neural receptive field emergence in scale-free networks. Studies how receptive fields organize and couple with stimulus space embedding without fine-tuning. Activation: receptive field geometry, scale-free networks, stimulus space embedding, neural geometry.

Neural Receptive Fields and Effective Geometry of Scale-Free Networks

Geometric framework explaining how receptive fields emerge and organize within scale-free brain networks, and how neural dynamics couple with stimulus space embedding without requiring connectivity fine-tuning.

Metadata

• Source: arXiv:2509.25453
• Authors: Vasilii Tiselko, Alexander Gorsky, Yuri Dabaghian
• Published: 2025-09

Core Methodology

Key Innovation

Demonstrates that receptive field organization emerges naturally from the effective geometry of scale-free network topology, eliminating the need for fine-tuned connectivity parameters typically required in receptive field models.

Technical Framework

1. Scale-Free Network Model: Use power-law degree distribution network as the substrate
2. Effective Geometry: Derive an effective geometric space from the network topology
3. Receptive Field Emergence: Show that neurons naturally develop receptive fields aligned with the effective geometry
4. Stimulus Space Coupling: Demonstrate how neural dynamics automatically couple to stimulus space through the geometric structure
5. Analysis: Characterize receptive field properties (size, shape, overlap) as functions of network topology

Theoretical Foundation

• Scale-free networks naturally embed in hyperbolic or curved metric spaces
• The effective geometry constrains how signals propagate through the network
• Receptive fields emerge as geometric "windows" into stimulus space

Implementation Guide

Prerequisites

• Network topology data (or synthetic scale-free network generation)
• Geometric embedding tools
• Stimulus representation framework

Step-by-Step

1. Generate or obtain scale-free network (power-law degree distribution)
2. Compute effective geometry (hyperbolic embedding or diffusion distance)
3. Assign stimulus positions in embedding space
4. Simulate neural dynamics on the network
5. Extract receptive fields from neural responses
6. Analyze geometric properties of receptive field organization

Code Example

import networkx as nx
import numpy as np

def generate_scalefree_network(n_nodes, m_edges=2):
    """Generate scale-free network using preferential attachment."""
    G = nx.barabasi_albert_graph(n_nodes, m_edges)
    return G

def compute_diffusion_geometry(adj_matrix, t_steps=10):
    """Compute diffusion-based effective geometry."""
    D = np.diag(adj_matrix.sum(axis=1))
    L = D - adj_matrix
    dist = np.linalg.matrix_exp(-L * t_steps)
    return dist

Applications

• Understanding receptive field development in visual cortex
• Modeling sensory processing in brain-like networks
• Neuromorphic vision system design
• Network-based feature learning

Pitfalls

• Scale-free assumption may not hold for all brain regions
• Effective geometry is an approximation of true neural dynamics
• Stimulus space coupling depends on appropriate embedding choice

Related Skills

• brain-inspired-attention-mechanisms
• primary-visual-cortex-v1-functions
• non-euclidean-visual-space-information-geometry