ElementMerc/qwen3-1.7b-abliterated
Abliterated vs Qwen/Qwen3-1.7B. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj) changed, in 13 of 28 layers. Abliteration edits exactly these matrices to remove a "refusal direction"; everything else is untouched.
Tensors changed: Embeddings 0/1, Norms 0/113, MLP 13/84, Attention 13/112
2 derivatives checked
- mradermacher/qwen3-1.7b-abliterated-i1-GGUF: Quantization
- ElementMerc/qwen3-1.7b-abliterated-GGUF: Same weights, converted