README.md

metadata

title: README
emoji: 👀
colorFrom: gray
colorTo: indigo
sdk: static
pinned: false

The official organization of the Heretic project

Heretic is a tool that removes censorship (aka "safety alignment") from transformer-based language models without expensive post-training. It combines an advanced implementation of directional ablation, also known as "abliteration" (Arditi et al. 2024, Lai 2025 (1, 2)), with a TPE-based parameter optimizer powered by Optuna.

The purpose of this organization is to publish and curate high-quality abliterated models made using Heretic.

Membership is open to anyone who has either

1. contributed code to the Heretic project, or
2. published a well-received model made using Heretic.

To become a member, please join our Discord and request an invitation.