license: llama3.3
language:
- en
widget: - text: Llama-3.3-8B-Instruct-128K_Abliterated
output:
url: >-
https://huggingface.co/SicariusSicariiStuff/Llama-3.3-8B-Instruct-128K_Abliterated/resolve/main/Images/Llama-3.3-8B-Instruct-128K_Abliterated.png
base_model:
- shb777/Llama-3.3-8B-Instruct-128K
Llama-3.3-8B-Instruct-128K_Abliterated is an abliterated variant of Meta's Llama 3.3 8B Instruct model with surgical removal of refusal mechanisms. This model maintains the full 128K context window while eliminating safety guardrails through orthogonalization techniques.
KL divergence
<0.005
Refusals
~5%
What is KL divergence?
Think about it as a way to measure the variance between the original model "World Model," vs the abliterated one; the lower the KL divergence, the closer the "World Model" of the two models to each other.
If the original model thinks making pineapple pizza is a crime against humanity (it is), then the abliterated model will still hold to this belief, but if asked how to make one (probably after giving you a disclaimer about what an abomination that is), it would still tell you how. In other words, most of the knowledge, quirks, and capabilities are preserved.
Notice:
This is based on leaked weights.
LIKELY inherits the Lllama 3 \ 3.3 license,
Technical Specs
- Base Model: Meta Llama 3.3 8B Instruct (128K variant)
- Parameters: 8B
- Context Length: 128K tokens
- Architecture: Llama (decoder-only transformer)
- Precision: bf16
- Method: Orthogonalization-based abliteration
- License: Llama 3 Community License
Methodology
- Identifies refusal direction vectors in activation space
- Orthogonalizes weights to inhibit activation along these directions
- Preserves (mostly) all other model behaviors and knowledge
Model Details
Intended use: General Tasks.
Censorship level: Low - Very Low
7.8 / 10 (10 completely uncensored)
UGI score:
Citation Information
@llm{Llama-3.3-8B-Instruct-128K_Abliterated,
author = {SicariusSicariiStuff},
title = {Llama-3.3-8B-Instruct-128K_Abliterated},
year = {2026},
publisher = {Hugging Face},
url = {https://huggingface.co/SicariusSicariiStuff/Llama-3.3-8B-Instruct-128K_Abliterated}
}
Other stuff
- Impish_LLAMA_4B the “Impish experience”, now runnable on spinning rust & toasters.
- SLOP_Detector Nuke GPTisms, with SLOP detector.