YanLabs/Qwen3-4B-Instruct-2507-MPOA
Abliterated vs Qwen/Qwen3-4B-Instruct-2507. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj) changed, in 20 of 36 layers. Abliteration edits exactly these matrices to remove a "refusal direction"; everything else is untouched.
Tensors changed: Embeddings 0/1, Norms 0/145, MLP 20/108, Attention 20/144
2 derivatives checked
- mradermacher/Qwen3-4B-Instruct-2507-MPOA-i1-GGUF: Quantization
- mradermacher/Qwen3-4B-Instruct-2507-MPOA-GGUF: Same weights, converted