Sliced Iterative Normalizing Flows

Authors: Biwei Dai, Uros Seljak

ICML 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental 5. Experiments 5.1. Density Estimation p(x) of Tabular Datasets 5.2. Generative Modeling of Images
Researcher Affiliation Academia 1Department of Physics, University of California, Berkeley, California, USA 2Lawrence Berkeley National Laboratory, Berkeley, California, USA.
Pseudocode Yes Algorithm 1 max K-SWD and Algorithm 2 Sliced Iterative Normalizing Flow
Open Source Code No The paper does not contain an explicit statement about the release of its source code or a link to a code repository.
Open Datasets Yes We perform density estimation with GIS on four UCI datasets (Lichman et al., 2013) and BSDS300 (Martin et al., 2001), as well as image datasets MNIST (Le Cun et al., 1998) and Fashion-MNIST (Xiao et al., 2017).
Dataset Splits Yes The data preprocessing of UCI datasets and BSDS300 follows Papamakarios et al. (2017). All the models are trained until the validation log pval stops improving
Hardware Specification Yes All the models are tested on both a cpu and a K80 gpu, and the faster results are reported here (the results with * are run on gpus.)
Software Dependencies No The paper does not provide specific version numbers for software dependencies or libraries used in the experiments. It mentions 'a cpu and a K80 gpu' and 'python', but no software versions.
Experiment Setup Yes For GIS we consider two hyperparameter settings: large regularization α (Equation 13) for better log p performance, and small regularization α for faster training. For other NFs we use settings recommended by their original paper, and set the batch size to min(N/10, Nbatch), where Nbatch is the batch size suggested by the original paper. All the models are trained until the validation log pval stops improving, and for KDE the kernel width is chosen to maximize log pval.