library_name: transformers
tags: []
Llama-3-8B-Instruct-abliterated-v2 Model Card
This is meta-llama/Llama-3-8B-Instruct with orthogonalized bfloat16 safetensor weights, generated with the methodology that was described in the preview paper/blog post: 'Refusal in LLMs is mediated by a single direction' which I encourage you to read to understand more.
TL;DR: this model has had certain weights manipulated to "inhibit" the model's ability to express refusal. It is not in anyway guaranteed that it won't refuse you, understand your request, it may still lecture you about ethics/safety, etc. It is tuned in all other respects the same as the original 8B instruct model was, just with the strongest refusal direction orthogonalized out.
V2 statement
This model was trained with more data to better pinpoint the "refusal direction".
This model is MUCH better at directly and succinctly answering requests without producing even so much as disclaimers.
Quirkiness awareness notice
This model may come with interesting quirks, as I obviously haven't extensively tested it, and the methodology being so new. I encourage you to play with the model, and post any quirks you notice in the community tab, as that'll help us further understand what this orthogonalization has in the way of side effects. The code I used to generate it (and my published 'Kappa-3' model which is just Phi-3 with the same methodology applied) is available in the Python notebook ortho_cookbook.ipynb.
If you manage to develop further improvements, please share! This is really the most primitive way to use ablation, but there are other possibilities that I believe are as-yet unexplored.