mjalali/renyi-kernel-entropy

[NeurIPS 2023] Code base for the Renyi Kernel Entropy (RKE) metric for generative models.

23
/ 100
Experimental

This tool helps machine learning researchers and practitioners rigorously evaluate how well their generative models capture the full diversity of real-world data, especially when that data has many distinct 'modes' or clusters, such as different categories of images. You input features extracted from both real data and your model's generated data, and it outputs a score that quantifies the number of modes your model has successfully learned and generated. It's designed for researchers developing or comparing new generative AI models.

No commits in the last 6 months.

Use this if you are a machine learning researcher or engineer developing generative AI models and need a reliable metric to quantify how many distinct data patterns your models produce compared to real-world data.

Not ideal if you are primarily interested in the visual quality or fidelity of generated samples, or if your data does not have a clearly multi-modal structure.

generative-ai model-evaluation image-synthesis machine-learning-research data-diversity
Stale 6m No Package No Dependents
Maintenance 2 / 25
Adoption 5 / 25
Maturity 16 / 25
Community 0 / 25

How are scores calculated?

Stars

13

Forks

Language

Python

License

MIT

Last pushed

Jun 18, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/generative-ai/mjalali/renyi-kernel-entropy"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.