Rate-Based RNN Examples
=======================

These examples focus on rate-based recurrent models and comparisons with BPTT.
The runnable scripts live in the repository's
`examples directory <https://github.com/chaobrain/braintrace/tree/main/examples>`__.

* `100-gru-on-copying-task.py <https://github.com/chaobrain/braintrace/blob/main/examples/100-gru-on-copying-task.py>`__
  trains a GRU on the copying task and compares online diagonal RTRL with BPTT.
  API: :doc:`GRUCell </apis/generated/braintrace.nn.GRUCell>`.
* `101-integrator-rnn.py <https://github.com/chaobrain/braintrace/blob/main/examples/101-integrator-rnn.py>`__
  trains a ``MiniGRU`` integrator and plots both predictions and training
  losses. API: :doc:`MiniGRU </apis/generated/braintrace.nn.MiniGRU>`.

The comparisons demonstrate behavior on the stated model and task. They do
not establish general gradient equivalence, because ``D_RTRL`` retains a
diagonal hidden-Jacobian approximation.

Begin with the :doc:`RNN Online Learning </tutorials/rnn_online_learning>`
workflow for the complete training sequence.

Related API
-----------

* :doc:`D-RTRL </apis/generated/braintrace.D_RTRL>`
* :doc:`SnAp </apis/generated/braintrace.SnAp>`
* :doc:`Neural Network Layers </apis/nn>`