Expand description
Activation function kernels for neural networks
Implements common activation functions (ReLU, Sigmoid, etc.)
Structs§
- Gelu
Kernel - GELU (Gaussian Error Linear Unit) activation function kernel Used heavily in modern transformer models and neural networks
- Leaky
Relu Kernel - LeakyReLU activation function kernel LeakyReLU(x) = max(α*x, x) where α is typically 0.01
- Relu
Kernel - ReLU activation function kernel
- Sigmoid
Kernel - Sigmoid activation function kernel
- Swish
Kernel - Swish (SiLU) activation function kernel Swish(x) = x * sigmoid(x) = x / (1 + exp(-x))
- Tanh
Kernel - Tanh activation function kernel