AI RESEARCH
A Constant-Time Implementation Methodology for Activation Functions on Microcontrollers
arXiv CS.AI
•
ArXi:2605.22441v1 Announce Type: cross Embedded neural-network inference can leak information through timing side channels, including leakage caused by the evaluation of activation functions. This work proposes a constant-time implementation methodology for activation functions on embedded microcontrollers and validates it on ReLU, sigmoid, tanh, GELU, and Swish on an ARM Cortex-M4 platform.