Gemma 3 abliterated

This is Gemma 3 4b with [abliteration technique] (https://www.lesswrong.com/posts/jGuXSZgv6qfdhMCuJ/refusal-in-llms-is-mediated-by-a-single-direction) applied to reduce refusals. (with high scale_factor=2.0 as it was very stubborn originally)

Tested its work in VLLM. It does give the disclaimers sometimes, but more or less often executes the instructions as needed, and the disclaimers are less aggressive.

Downloads last month
29
Safetensors
Model size
4.3B params
Tensor type
BF16
·
FP16
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for kabachuha/gemma3-4b-it-abliterated

Finetuned
(104)
this model
Quantizations
2 models