Learning Important Features Through Propagating Activation Differences

Authors: Avanti Shrikumar, Peyton Greenside, Anshul Kundaje

ICML 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We apply Deep LIFT to models trained on MNIST and simulated genomic data, and show significant advantages over gradient-based methods. (Abstract)
Researcher Affiliation Academia 1Stanford University, Stanford, California, USA.
Pseudocode No The paper describes its rules and method using mathematical formulations but does not contain any structured pseudocode or algorithm blocks.
Open Source Code Yes Video tutorial: http://goo.gl/qKb7pL, code: http://goo.gl/RM8jvH.
Open Datasets Yes We train a convolutional neural network on MNIST (Le Cun et al., 1999) using Keras (Chollet, 2015) to perform digit classification and obtain 99.2% test-set accuracy.
Dataset Splits No The paper mentions training on MNIST and evaluating on its 'test-set', but it does not explicitly specify train/validation/test dataset splits, percentages, or sample counts needed to reproduce the partitioning.
Hardware Specification No The paper does not specify any particular hardware used for the experiments, such as GPU models, CPU models, or memory details.
Software Dependencies No The paper mentions using 'Keras (Chollet, 2015)' but does not specify a version number for Keras or any other software dependencies.
Experiment Setup Yes The architecture consists of two convolutional layers, followed by a fully connected layer, followed by the softmax output layer (see Appendix D for full details on model architecture and training). (Section 4.1) Details of the simulation, network architecture and predictive performance are given in Appendix F. (Section 4.2)