prithivMLmods/Qwen3.5-9B-Unredacted-MAX
Abliterated vs Qwen/Qwen3.5-9B. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj, out_proj) changed, in 31 of 32 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 31/208, Attention 31/332, Norms 0/215
4 derivatives checked
- mradermacher/Qwen3.5-9B-Unredacted-MAX-i1-GGUF: Quantization
- mradermacher/Qwen3.5-9B-Unredacted-MAX-GGUF: Same weights, converted
- prithivMLmods/Q3.5-9B-DS-v4-Flash-DA: LoRA-style
- prithivMLmods/Q3.5-9B-GLM-5.1-DA: LoRA-style