zebulon-prime/Qwen3.8-27B-Dominatrix-abliterated
Abliterated vs allura-org/Qwen3.8-27B-Dominatrix. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj, out_proj) changed, in 49 of 64 layers. Abliteration edits exactly these matrices to remove a "refusal direction"; everything else is untouched.
Tensors changed: lm_head 0/1, Embeddings 0/5, Norms 0/325, Attention 49/560, MLP 49/307, other 0/1
2 derivatives checked
- mradermacher/Qwen3.8-27B-Dominatrix-abliterated-i1-GGUF: Quantization
- mradermacher/Qwen3.8-27B-Dominatrix-abliterated-GGUF: Quantization