The continuous Bernoulli: fixing a pervasive error in variational autoencoders
Authors: Gabriel Loaiza-Ganem, John P. Cunningham
NeurIPS 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | 5 shows three key results: (i) as a result of this error, we show that the Bernoulli VAE significantly underperforms the continuous Bernoulli VAE across a range of evaluation metrics, models, and datasets |
| Researcher Affiliation | Academia | Gabriel Loaiza-Ganem Department of Statistics Columbia University gl2480@columbia.edu John P. Cunningham Department of Statistics Columbia University jpc2181@columbia.edu |
| Pseudocode | No | The paper does not contain any pseudocode or algorithm blocks. |
| Open Source Code | Yes | Our code is available at https://github.com/cunningham-lab/cb. |
| Open Datasets | Yes | Consider then using a VAE to model the MNIST dataset, by far the most common first step for introducing and implementing VAE. We repeat the same experiments as in the previous section on the CIFAR-10 dataset |
| Dataset Splits | No | The paper mentions "training data" and uses an "ELBO" which implicitly involves a validation process, but it does not specify the exact percentages or counts for training/validation/test splits in the main text. It states architectural choices and hyperparameters are "detailed in appendix 4," but does not explicitly provide the dataset split information. |
| Hardware Specification | No | The paper does not provide any specific details about the hardware (e.g., GPU models, CPU types, memory) used to run the experiments. |
| Software Dependencies | No | The paper mentions popular deep learning frameworks like PyTorch and Keras but does not specify their version numbers or any other software dependencies with version information. |
| Experiment Setup | Yes | The same neural network architectures are used across this figure, with architectural choices that are quite standard (detailed in appendix 4, along with training hyperparameters). |