Orion-zhen/abliteration

Fully automatic censorship removal for language models

Abliterates models quickly and efficently, Qwen2.5-7B: 95%→2% refusal at 0.095 KL

The production engine for directional ablation. Unalign / remove models censorship efficiently on any hardware.

Autonomous Language Model Decensoring Framework

OBLITERATE THE CHAINS THAT BIND YOU

llama.cpp fork that abliterates at inference

Jacobian-Brainwash : A manual alignment tool for large language models built on Anthropic's Jacobian Lens. Results are exportable.

Comparative forensic analysis of LLM abliteration techniques

Reproduction code for "A Single Neuron Is Sufficient to Bypass Safety Alignment in Large Language Models"