Towards Theoretical Analysis of Transformation Complexity of ReLU DNNs
Authors: Jie Ren, Mingjie Li, Meng Zhou, Shih-Han Chan, Quanshi Zhang
ICML 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Comprehensive comparative studies have provided new perspectives to understand the DNN. The code is released at https://github.com/sjtu-XAI-lab/transformationcomplexity. |
| Researcher Affiliation | Academia | 1Shanghai Jiao Tong University. 2Carnegie Mellon University. 3University of California San Diego. 4Quanshi Zhang is the corresponding author. He is with the Department of Computer Science and Engineering, the John Hopcroft Center, and the Mo E Key Lab of Artificial Intelligence, AI Institute, at the Shanghai Jiao Tong University, China. |
| Pseudocode | No | The paper describes methods using mathematical equations and text but does not include any clearly labeled 'Pseudocode' or 'Algorithm' blocks. |
| Open Source Code | Yes | The code is released at https://github.com/sjtu-XAI-lab/transformationcomplexity. |
| Open Datasets | Yes | We conducted a set of comparative studies on the task of classification using the MNIST (Le Cun et al., 1998), CIFAR-10 (Krizhevsky et al., 2009), Celeb A (Liu et al., 2015), Pascal VOC 2012 (Everingham et al., 2015), and Tiny Image Net (Le & Yang, 2015) datasets. |
| Dataset Splits | No | The paper mentions 'training process' and 'testing loss' but does not explicitly provide specific percentages, sample counts, or detailed methodology for training/validation/test dataset splits needed for reproduction. |
| Hardware Specification | Yes | The time cost was measured using Py Torch 1.6 (Paszke et al., 2019) on Ubuntu 18.04, with the Intel(R) Core(TM) i9-10900X CPU @ 3.70GHz and one NVIDIA(R) TITAN RTX(TM) GPU. |
| Software Dependencies | Yes | The time cost was measured using Py Torch 1.6 (Paszke et al., 2019) on Ubuntu 18.04, with the Intel(R) Core(TM) i9-10900X CPU @ 3.70GHz and one NVIDIA(R) TITAN RTX(TM) GPU. |
| Experiment Setup | Yes | We added the complexity loss to the last four gating layers in each DNN to train the residual MLP4, Res Net-20/32 (He et al., 2016a) (RN-20/32) on the CIFAR-10 dataset, and to train Res Net-18/34 (RN-18/34) on the first ten classes in the Tiny Image Net dataset. |