3 ms·
This seems analogous to the weights in a neural network. In training, essential information about the training set is stored in weights and the rest is discarde
by shrimpx 5y ago
This seems analogous to the weights in a neural network. In training, essential information about the training set is stored in weights and the rest is discarded. You can’t recover a training sample from a trained network.