MuXodious/Qwen3.8-27B-absolute-heresy
Abliterated vs Qwen/Qwen3.8-27B. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj, out_proj) changed, in 33 of 64 layers. Abliteration edits exactly these matrices to remove a "refusal direction"; everything else is untouched.
Tensors changed: Norms 0/325, Attention 24/560, MLP 9/307, Embeddings 0/5, lm_head 0/1, other 0/1
5 derivatives checked
- mradermacher/Qwen3.8-27B-absolute-heresy-i1-GGUF: Quantization
- mradermacher/Qwen3.8-27B-absolute-heresy-GGUF: Quantization
- Starw1/Qwen3.8-27B-absolute-heresy-W4A16: Quantization
- MuXodious/Qwen3.8-27B-absolute-heresy-GGUF: Quantization
- MuXodious/Qwen3.8-27B-absolute-heresy-mlx-4Bit: Quantization