<p>About the part of prediction, if I could access underlying network, it would be easier</p>

<pre><code>using ActivateFunction = typename HiddenLayer::ActivateFunction;

arma::mat const encodeInput = w1 * input+ arma::repmat(b1, 1, input.n_cols);
ActivateFunction::fn(encodeInput , features);
</code></pre>

<p>As you can see, to extract the features from the input, I do not need to forward the whole layers, I would need to access underlying network of the FFN, could I add a public api to access the network?</p>

<p style="font-size:small;-webkit-text-size-adjust:none;color:#666;">&mdash;<br>Reply to this email directly or <a href="https://github.com/mlpack/mlpack/pull/451#issuecomment-164171021">view it on GitHub</a>.<img alt="" height="1" src="https://github.com/notifications/beacon/AJ4bFMpUdgm-IV3QEZTarYNlI-iZDxl2ks5pPFRAgaJpZM4GAqt4.gif" width="1" /></p>
<div itemscope itemtype="http://schema.org/EmailMessage">
<div itemprop="action" itemscope itemtype="http://schema.org/ViewAction">
  <link itemprop="url" href="https://github.com/mlpack/mlpack/pull/451#issuecomment-164171021"></link>
  <meta itemprop="name" content="View Pull Request"></meta>
</div>
<meta itemprop="description" content="View this Pull Request on GitHub"></meta>
</div>