Invertible DenseNets
This work addresses the need for more parameter-efficient invertible networks in machine learning, though it appears incremental as it builds on existing DenseNet and flow-based methods.
The paper tackled the problem of parameter inefficiency in invertible neural networks by introducing Invertible Dense Networks (i-DenseNets), which outperformed Residual Flows in negative log-likelihood on toy, MNIST, and CIFAR10 datasets under equal parameter budgets.
We introduce Invertible Dense Networks (i-DenseNets), a more parameter efficient alternative to Residual Flows. The method relies on an analysis of the Lipschitz continuity of the concatenation in DenseNets, where we enforce the invertibility of the network by satisfying the Lipschitz constraint. Additionally, we extend this method by proposing a learnable concatenation, which not only improves the model performance but also indicates the importance of the concatenated representation. We demonstrate the performance of i-DenseNets and Residual Flows on toy, MNIST, and CIFAR10 data. Both i-DenseNets outperform Residual Flows evaluated in negative log-likelihood, on all considered datasets under an equal parameter budget.