A Game Theoretic Analysis of Additive Adversarial Attacks and Defenses

Authors: Ambar Pal, Rene Vidal

NeurIPS 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Our experiments are conducted on the MNIST and FMNIST datasets restricted to two classes. We train a 4-layer convolutional neural network with ReLU activation functions for this binary classification task. The classification results are shown in Table 1, from which we can draw two main conclusions:
Researcher Affiliation Academia Ambar Pal Mathematical Institute for Data Science Johns Hopkins University ambar@jhu.edu René Vidal Mathematical Institute for Data Science Johns Hopkins University rvidal@jhu.edu
Pseudocode No No pseudocode or algorithm blocks were found in the paper's main text.
Open Source Code No The paper does not provide an explicit statement about releasing source code or a link to a code repository for the described methodology.
Open Datasets Yes Our experiments are conducted on the MNIST and FMNIST datasets restricted to two classes.
Dataset Splits No The paper mentions using a 'finite training set' and discusses 'generalization bounds', but does not explicitly specify the training, validation, or test dataset splits (e.g., percentages or sample counts) needed for reproduction.
Hardware Specification No No specific hardware (e.g., GPU/CPU models, memory details) used for running the experiments was mentioned in the paper.
Software Dependencies No The paper mentions training a '4-layer convolutional neural network with ReLU activation functions' but does not specify any software dependencies (e.g., PyTorch, TensorFlow) with version numbers.
Experiment Setup No The paper states 'A detailed description, as well as more experimental details can be found in Sec. D of the Appendix.', indicating that some details exist elsewhere in the paper, but the provided text itself does not contain specific experimental setup details such as hyperparameters or training configurations.