q-Neurons: stochastic Jackson-derivative activations consistently beat standard ones

FrnkNlsn · x · 2026-09-23

A paper by Frank Nielsen and Ke Sun (IEEE TNNLS 2020) introduces q-neurons, neurons whose activations are based on Jackson's q-derivatives with stochastic parameter q. The authors show the generalization is scalable and easy to implement, with experiments demonstrating consistently improved training and test losses over state-of-the-art standard activation functions.

Original post →

More from Research

Research channel →