pnnl/modelsym
Code to run the experiments of the Neurips 2022 paper On the Symmetries of Deep Learning Models and their Internal Representations.
This project provides code to investigate the fundamental symmetries within deep learning models and how they represent information internally. It takes trained neural networks and uses advanced techniques like 'stitching' layers and 'alignment' metrics to analyze and visualize their internal structure and behavior. This is designed for machine learning researchers and academics exploring the theoretical underpinnings of neural networks.
No commits in the last 6 months.
Use this if you are a researcher studying the theoretical aspects of deep learning, specifically focusing on model symmetries and internal representations.
Not ideal if you are looking to apply deep learning models for practical tasks like image classification or natural language processing without a focus on their mathematical properties.
Stars
7
Forks
3
Language
Jupyter Notebook
License
—
Category
Last pushed
Nov 08, 2022
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/pnnl/modelsym"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
pyg-team/pytorch_geometric
Graph Neural Network Library for PyTorch
a-r-j/graphein
Protein Graph Library
raamana/graynet
Subject-wise networks from structural MRI, both vertex- and voxel-wise features (thickness, GM...
pykale/pykale
Knowledge-Aware machine LEarning (KALE): accessible machine learning from multiple sources for...
dmlc/dgl
Python package built to ease deep learning on graph, on top of existing DL frameworks.