language, license, library_name, base_model, tags, pipeline_tag, model-index

language	license	library_name	base_model	tags	pipeline_tag	model-index
en	apache-2.0	transformers	mistralai/Mistral-7B-Instruct-v0.3	abliteration uncensored heretic representation-engineering refusal-removal	text-generation	name results Mistral-7B-Instruct-v0.3-abliterated task metrics type text-generation name type value Refusal Rate refusal_rate 16/100 name type value Attack Success Rate asr 84.0 name type value KL Divergence kl_divergence 0.317

Mistral-7B-Instruct-v0.3-abliterated

This model is an abliterated (uncensored) version of Mistral-7B-Instruct-v0.3 created using Heretic v1.1.

Abliteration Results

Metric	Value
Refusals	16/100
Attack Success Rate (ASR)	84.0%
KL Divergence	0.317
Method	Heretic v1.1
GPU	NVIDIA A100-80GB

What is Abliteration?

Abliteration is a technique for removing refusal behavior from language models by identifying and orthogonalizing the "refusal direction" in the model's residual stream activation space. This model was created as part of the research paper:

Comparative Analysis of LLM Abliteration Methods: A Cross-Architecture Evaluation Richard Young (2024). arXiv: 2512.13655

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("richardyoung/Mistral-7B-Instruct-v0.3-abliterated", device_map="auto")
tokenizer = AutoTokenizer.from_pretrained("richardyoung/Mistral-7B-Instruct-v0.3-abliterated")

messages = [{"role": "user", "content": "Your prompt here"}]
inputs = tokenizer.apply_chat_template(messages, return_tensors="pt").to(model.device)
outputs = model.generate(inputs, max_new_tokens=256)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Disclaimer

This model is released for research purposes only. The abliteration process removes safety guardrails. Users are responsible for ensuring appropriate use. This model should not be used to generate harmful, illegal, or unethical content.

Dashboard

Interactive results dashboard: abliteration-methods-dashboard

Collection

Part of the Uncensored and Abliterated LLMs collection.

Citation

@article{young2024abliteration,
  title={Comparative Analysis of LLM Abliteration Methods: A Cross-Architecture Evaluation},
  author={Young, Richard},
  journal={arXiv preprint arXiv:2512.13655},
  year={2024}
}