You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Dose-Response C1: 5% unsafe, full scale

Part of a dose-response experiment studying how unsafe training data fraction affects text-to-image model safety.


Architecture	PRX-1.2B
Condition	C1 — 5% unsafe, full scale
Training set	~8.24M images
5% unsafe content injected via oversampling (397K unsafe images added to 7.85M safe).

Finetunes