Von Neumann Networks offer parameter-efficient AI, outperforming deep learning variants

Researchers have introduced a new type of artificial neuron, termed the Von Neumann neuron, inspired by John von Neumann's mid-twentieth-century computational model. These neurons, when organized into Von Neumann Networks (VNNs), can learn specialized roles and self-engineer their architecture based on input and output configurations. The VNN framework extends neural operators and learns Green's functions through convolutions on a cellular topology, demonstrating superior performance and parameter efficiency compared to traditional deep learning models on basic tasks. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Introduces a novel neural network architecture inspired by historical computational models, potentially offering new avenues for efficient learning and task generalization.

RANK_REASON Academic paper introducing a novel neural network architecture and neuron type. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• John von Neumann
• Von Neumann Networks
• Von Neumann neuron
• Green's functions
• Cellular Machines
• deep learning
• multi-layered perceptrons

• paper
• other