Learning Compact Boolean Networks

Shengpu Wang, Yuhao Mao, Yani Zhang +1

Floating-point neural networks dominate modern machine learning but incur substantial inference costs, motivating emerging interest in Boolean networks for resource-constrained dep…