toandev/Gemma4-12B-Uncensored
Abliterated vs google/gemma-4-12B-it. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj) changed, in 24 of 48 layers. Abliteration edits exactly these matrices to remove a "refusal direction"; everything else is untouched.
Tensors changed: Embeddings 0/12, Norms 0/289, other 0/48, MLP 24/144, Attention 24/184
3 derivatives checked
- mradermacher/Gemma4-12B-Uncensored-i1-GGUF: Quantization
- mradermacher/Gemma4-12B-Uncensored-GGUF: Quantization
- toandev/Gemma4-12B-Uncensored-GGUF: Quantization