From d9243a78151b3dc23ce71a2b47e47cf319d5f369 Mon Sep 17 00:00:00 2001
From: Ruiqing Yan <YanLabs@users.noreply.huggingface.co>
Date: Sat, 17 Jan 2026 09:11:12 +0000
Subject: [PATCH] Update README.md

---
 README.md | 63 +++++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 63 insertions(+)

diff --git a/README.md b/README.md
index 26dcb40..f5dfe87 100644
--- a/README.md
+++ b/README.md
@@ -1,2 +1,65 @@
+---
+license: apache-2.0
+base_model:
+- YanLabs/Llama-3.3-8B-Instruct-MPOA
+pipeline_tag: text-generation
+---
+
+# YanLabs/Llama-3.3-8B-Instruct-MPOA
+
 This is an abliterated version of shb777/Llama-3.3-8B-Instruct (originally allura-forge/Llama-3.3-8B-Instruct).
 Recommended temp >=1.0
+
+**⚠️ Warning**: Safety guardrails and refusal mechanisms have been removed through abliteration. This model may generate harmful content and is intended for mechanistic interpretability research only.
+
+## Model Details
+
+### Model Description
+
+This model applies **norm-preserving biprojected abliteration** to remove refusal behaviors while preserving the model's original capabilities. The technique surgically removes "refusal directions" from the model's activation space without traditional fine-tuning.
+
+- **Developed by**: YanLabs
+- **Model type**: Causal Language Model (Transformer)
+- **License**: apache-2.0
+- **Base model**:  [shb777/Llama-3.3-8B-Instruct-128K](https://huggingface.co/shb777/Llama-3.3-8B-Instruct-128K)
+
+### Model Sources
+
+- **Base Model**:  [shb777/Llama-3.3-8B-Instruct-128K](https://huggingface.co/shb777/Llama-3.3-8B-Instruct-128K)
+- **Abliteration Tool**: [jim-plus/llm-abliteration](https://github.com/jim-plus/llm-abliteration)
+- **Paper**: [Norm-Preserving Biprojected Abliteration](https://huggingface.co/blog/grimjim/norm-preserving-biprojected-abliteration)
+
+## Uses
+
+### Intended Use
+
+- **Research**: Mechanistic interpretability studies
+- **Analysis**: Understanding LLM safety mechanisms
+- **Development**: Testing abliteration techniques
+
+### Out-of-Scope Use
+
+- ❌ Production deployments
+- ❌ User-facing applications
+- ❌ Generating harmful content for malicious purposes
+
+## Limitations
+
+- Abliteration does not guarantee complete removal of all refusals
+- May generate unsafe or harmful content
+- Model behavior may be unpredictable in edge cases
+- No explicit harm prevention mechanisms remain
+
+## Citation
+
+If you use this model in your research, please cite:
+
+```bibtex
+@misc{lama-3.3-8B-Instruct-MPOA,
+  author = {YanLabs},
+  title = {lama-3.3-8B-Instruct-MPOA},
+  year = {2025},
+  publisher = {HuggingFace},
+  howpublished = {\url{https://huggingface.co/YanLabs/Llama-3.3-8B-Instruct-MPOA}},
+  note = {Abliterated using norm-preserving biprojected technique}
+}
\ No newline at end of file