wangzhang/gemma-4-12B-it-abliterix
Abliterated vs google/gemma-4-12B-it. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj) changed, in 31 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 31/144, Attention 27/184
2 derivatives checked
- mradermacher/gemma-4-12B-it-abliterix-i1-GGUF: Quantization
- mradermacher/gemma-4-12B-it-abliterix-GGUF: Quantization