llmfan46/Qwen3.5-27B-ultra-uncensored-heretic-v2
Abliterated vs Qwen/Qwen3.5-27B. Looks like abliteration (refusal removal): only the output projections (o_proj, out_proj) changed, in 28 of 64 layers. Abliteration edits exactly these matrices to remove a "refusal direction"; everything else is untouched.
Tensors changed: lm_head 0/1, Embeddings 0/4, MLP 0/304, Attention 28/556, Norms 0/319
4 derivatives checked
- mradermacher/Qwen3.5-27B-ultra-uncensored-heretic-v2-i1-GGUF: Quantization
- llmfan46/Qwen3.5-27B-ultra-uncensored-heretic-v2-GGUF: Same weights, converted
- mradermacher/Qwen3.5-27B-ultra-uncensored-heretic-v2-GGUF: Quantization
- Jongsim/Qwen3.5-27B-heretic-v2-Opus-4.6-Distilled: LoRA-style
1 other repos hold byte-for-byte identical weight files.