ReLU
Domain: Math & Logic for AI
🎯 Motivation
Max(0, x). Simple, fast, and avoids the vanishing gradient problem for positive inputs.
📋 Concrete Example
ReLU(5)=5, ReLU(-3)=0. Gradient is 1 for positive, 0 for negative.
🔗 Analogy
Like a one-way valve - positive flow passes through unchanged, negative flow is blocked.