Building blocks
In nnodely, model behavior is defined through a collection of modular building blocks. These blocks implement elementary operations and structured transformations that can be combined to form complex neural models.
Building blocks operate on inputs, parameters, and intermediate signals, and represent the computational core of the architecture.
Each block encapsulates a specific mathematical or functional behavior and can be reused across different models.