SOURCE-LINKED INTELLIGENCE

NGN: Learning Neural Network Size as a Differentiable Count

arXiv · AI, language, vision and robotics · article · Sep 23, 2026 · UTC

Neural network size is usually chosen before training, separating architecture selection from weight optimization. We introduce the Neurogenesis Network (NGN), a differentiable parameterization for learning how many ordered structural components a model should use. For each ordered component group, one learnable boundary selects an active prefix while the model parameters are trained. The boundary can grow from a compact initialization and can be deployed by discarding components beyond the learned boundary. Controlled experiments examine convergence of the learned boundary, the performance of

Read original source ↗ Open in workspace

recordType

paper

region

Global