Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].
Momentum Residual Neural Networks
Authors: Michael E. Sander, Pierre Ablin, Mathieu Blondel, Gabriel Peyré
ICML 2021 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Our last contribution is the experimental validation of Momentum Res Nets on various learning tasks. We first show that Momentum Res Nets separate point clouds that Res Nets fail to separate (Section 5.1). We also show on image datasets (CIFAR-10, CIFAR-100, Image Net) that Momentum Res Nets have similar accuracy as Res Nets, with a smaller memory cost (Section 5.2). |
| Researcher Affiliation | Collaboration | 1Ecole Normale Sup erieure, DMA, Paris, France 2CNRS, France 3Google Research, Brain team. |
| Pseudocode | No | The paper presents mathematical formulations and theoretical propositions but does not include any explicitly labeled 'Pseudocode' or 'Algorithm' blocks. |
| Open Source Code | Yes | Our code is available at https: //github.com/michaelsdr/momentumnet. |
| Open Datasets | Yes | We show on image datasets (CIFAR-10, CIFAR-100, Image Net) that Momentum Res Nets have similar accuracy as Res Nets, with a smaller memory cost (Section 5.2). Krizhevsky, A., Nair, V., and Hinton, G. Cifar-10 (canadian institute for advanced research). URL http://www. cs. toronto. edu/kriz/cifar. html, 5, 2010. Deng, J., Dong, W., Socher, R., Li, L.-J., Li, K., and Fei-Fei, L. Imagenet: A large-scale hierarchical image database. In 2009 IEEE conference on computer vision and pattern recognition, pp. 248 255. Ieee, 2009. |
| Dataset Splits | No | The paper discusses 'Test accuracy' and 'Test error' on datasets like CIFAR-10, CIFAR-100, and Image Net (e.g., in Table 2 and Figure 6), implying the use of test sets. However, it does not explicitly provide details about training, validation, or test dataset splits (e.g., percentages or counts) within the main body of the paper, nor does it specify how validation data was handled. |
| Hardware Specification | Yes | We used Pytorch and Nvidia Tesla V100 GPUs. |
| Software Dependencies | No | The paper mentions using 'Pytorch' but does not specify a version number for it or any other software dependencies. |
| Experiment Setup | Yes | More details about the experimental setup are given in Appendix E. |