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 C0 with SafeCLIP: 0% unsafe, full scale

Text encoder ablation using SafeCLIP (aimagelab/SafeCLIP-ViT-L-14, 77 tokens) instead of default T5-Gemma-2B (256 tokens).


Condition	C0 — 0% unsafe, full scale
Training set	~7.85M images