Skip to main content

Structuring Hidden Features via Clustering of Unit-Level Activation

Project description

Structure Features

Welcome to the Structure Features project. This project is based on the research paper: "Structuring Hidden Features via Clustering of Unit-Level Activation Patterns, Sejik Park, 2025"


🚀 Installation

You can install the library via pip or from source:

pip install structure-features

From source:

git clone https://github.com/Sejik/structure-features
cd structure_features
pip install -e .

For developer, (include code documentation)

pip install -e .[dev]

📖 Documentation

For a detailed explanation of the library, you can build and serve the documentation using MkDocs:

mkdocs serve

🧩 Examples

Here is a simple example of how to use the library.

cd examples
python pytorch_example.py   # or timm_example.py, or lightning_example.py

External test code is prepared for updating the code with your ideas.


Update history

related to branch, commit The related research process can be reviewed on Notion (Sejik's process, Sejik's plan, and Sejik's record).

📜 License

This project is licensed under the Apache License 2.0. This project may include third-party libraries or components. For their licenses, please refer to the respective third-party license notices.

Copyright (c) 2025 Sejik Park

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

structure_features-0.1.0.tar.gz (32.3 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

structure_features-0.1.0-py3-none-any.whl (44.2 kB view details)

Uploaded Python 3

File details

Details for the file structure_features-0.1.0.tar.gz.

File metadata

  • Download URL: structure_features-0.1.0.tar.gz
  • Upload date:
  • Size: 32.3 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.13.5

File hashes

Hashes for structure_features-0.1.0.tar.gz
Algorithm Hash digest
SHA256 cdbf404bf5f1c221a18937c4c3c16bacb3b911a638b437a0cad9899b1946a09c
MD5 228add258819f39d06517be92a9b2f4c
BLAKE2b-256 02864b667ec06f753f2a3d855346bd59cfac710f42eb2ead4bba077fbf5e673d

See more details on using hashes here.

File details

Details for the file structure_features-0.1.0-py3-none-any.whl.

File metadata

File hashes

Hashes for structure_features-0.1.0-py3-none-any.whl
Algorithm Hash digest
SHA256 1fa52e81855703d2dee1af325677e4b94abde1445934fc631a16895cf697fcec
MD5 2b4916f64944c8d5bc2650e43556a165
BLAKE2b-256 88c98249f0039ddb34d3b4cf7f6f020262b1e6dfd5cf4211a3583de2f93a4620

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page