FlexNN 1
Fully connected neural network built from scratch with flexible n-layer design and multiple activations.
Loading...
Searching...
No Matches
activations Directory Reference
Directory dependency graph for activations:
include/activations

Files

 Activation.hpp
 Public activation enum and helpers for FlexNN.
 
 ActivationParameters.hpp
 Per-activation hyperparameters for FlexNN::Activations.
 
 detail.hpp
 Convenience re-export for activation detail headers.
 
 LeakyReLU.hpp
 Leaky ReLU with configurable alpha (default 0.01).
 
 None.hpp
 Linear (no) activation — identity forward/backward.
 
 ReLU.hpp
 ReLU activation — max(0, z) forward, (z>0) backward.
 
 Sigmoid.hpp
 Sigmoid activation — 1/(1+exp(-Z)) with clamping.
 
 Softmax.hpp
 Stable column-wise softmax.
 
 Tanh.hpp
 Tanh activation — std::tanh forward, (1-A^2) backward.