llmfan46/Qwen3.8-27B-Uncensored-Heretic-Native-MTP-Preserved
Abliterated vs Qwen/Qwen3.8-27B. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj, out_proj) changed, in 61 of 64 layers. Abliteration edits exactly these matrices to remove a "refusal direction"; everything else is untouched.
Tensors changed: Norms 0/325, Attention 61/560, MLP 38/307, Embeddings 0/5, lm_head 0/1, other 0/1
1 derivatives checked
- codavidgarcia/Qwen3.8-27B-Uncensored-Heretic-MTP-UD-GGUF: Quantization