ProtoVAE: A Trustworthy Self-Explainable Prototypical Variational Model
Authors: Srishti Gautam, Ahcène Boubekki, Stine Hansen, Suaiba Salahuddin, Robert Jenssen, Marina Höhne, Michael Kampffmeyer
NeurIPS 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In this section, we conduct extensive experiments to evaluate Proto VAE s trustworthiness, transparency, and ability to capture the diversity in the data. More specifically, we demonstrate the trustworthiness of our model in terms of predictive performance in Sec. 5.1. Qualitative evaluations are then conducted in Sec. 5.2 to verify the diversity and transparency properties, followed by a quantitative evaluation of the explanations corroborating its trustworthiness. |
| Researcher Affiliation | Academia | 1UiT The Arctic University of Norway 2Technical University of Berlin |
| Pseudocode | No | The paper describes the model architecture and loss functions using mathematical equations and textual descriptions, but does not provide pseudocode or algorithm blocks. |
| Open Source Code | Yes | Our code is available at https://github.com/Srishti Gautam/Proto VAE. |
| Open Datasets | Yes | We evaluate Proto VAE on 5 datasets, MNIST [26], Fashion MNIST [27] (f MNIST), CIFAR-10, [28], a subset of Quick Draw [29] and SVHN [30]. |
| Dataset Splits | No | The paper states that training details, including data splits, are reported in the supplementary material, but does not explicitly provide specific training/validation/test dataset splits in the main text. |
| Hardware Specification | No | The paper states that the total amount of compute and type of resources used are included in the supplementary material, but does not provide specific hardware details (e.g., GPU models, CPU types) in the main text. |
| Software Dependencies | No | The paper does not provide specific software dependencies with version numbers (e.g., Python version, library versions). |
| Experiment Setup | No | Further details about the datasets and additional implementation details, such as the detailed architecture and hyperparameters, are provided in the supplementary material Sec. S3 and S4. |