From 814884414c3a475b66bfd33d1edd0f5d520ffa9e Mon Sep 17 00:00:00 2001 From: Hanzo Dev Date: Fri, 27 Feb 2026 16:01:19 -0800 Subject: [PATCH] docs: update abliteration section to reference hanzoai/remove-refusals --- README.md | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/README.md b/README.md index 0037618..82f89c5 100644 --- a/README.md +++ b/README.md @@ -178,11 +178,11 @@ Part of the [Zen model family](https://zenlm.org) by [Hanzo AI](https://hanzo.ai Zen Pro 8B weights are derived from an abliterated base model. Abliteration removes refusal behavior by identifying and nullifying the "refusal direction" in the model's -residual stream, as described in -[Maxime Labonne's methodology](https://huggingface.co/blog/mlabonne/abliteration). +residual stream. **Method**: Directional ablation on the residual stream across all layers -**Source**: Zen MoDE (Mixture of Distilled Experts) architecture base +**Implementation**: [hanzoai/remove-refusals](https://github.com/hanzoai/remove-refusals) — Hanzo's production abliteration toolkit +**Technique**: [Refusal in LLMs is mediated by a single direction](https://www.lesswrong.com/posts/jGuXSZgv6qfdhMCuJ/refusal-in-llms-is-mediated-by-a-single-direction) — Arditi et al. **Effect**: Removes refusal behaviors while preserving all other capabilities **Identity layer**: Zen identity added via system prompt — full LoRA fine-tuning planned