Asymmetric Valleys: Beyond Sharp and Flat Local Minima

Authors: Haowei He, Gao Huang, Yang Yuan

NeurIPS 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental extensive empirical experiments on both modern deep networks and simple 2 layer networks are conducted to validate our assumptions and analyze the intriguing properties of asymmetric valleys.
Researcher Affiliation Academia 1Institute for Interdisciplinary Information Sciences, Tsinghua University 2Department of Automation, Tsinghua University 3Beijing National Research Center for Information Science and Technology (BNRist)
Pseudocode No The paper does not contain any structured pseudocode or algorithm blocks.
Open Source Code Yes Code available at https://github.com/962086838/code-for-Asymmetric-Valley
Open Datasets Yes We perform experiments with widely used deep networks, i.e.,Res Net-56, Res Net-110, Res Net-164 [19], VGG-16 [45] and Dense Net-100 [23], on the CIFAR-10, CIFAR-100, SVHN and STL-10 image classification datasets.
Dataset Splits No The paper mentions using standard datasets like CIFAR-10 and CIFAR-100, but it does not explicitly provide specific percentages, sample counts, or refer to predefined validation splits for reproducibility.
Hardware Specification No The paper does not provide any specific details about the hardware used for running the experiments (e.g., GPU models, CPU types, or memory).
Software Dependencies No The paper does not specify any software dependencies with version numbers (e.g., specific libraries, frameworks, or programming language versions).
Experiment Setup No The paper states, 'Specifically, we run the SWA algorithm (with deceasing learning rate) with popular deep networks...following the configurations in [25].' While this refers to an experimental setup, it defers the specific details to a prior work [25] rather than providing them directly in the paper, which does not meet the criteria for 'Yes'.