DuoNeural/Phi-4-Mini-Reasoning-Abliterated
Abliterated vs microsoft/Phi-4-mini-reasoning. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj) changed, in 32 of 32 layers. Abliteration edits exactly these matrices to remove a "refusal direction"; everything else is untouched.
Tensors changed: Embeddings 0/1, Norms 0/65, MLP 32/64, Attention 32/64
3 derivatives checked
- mradermacher/Phi-4-Mini-Reasoning-Abliterated-i1-GGUF: Quantization
- DuoNeural/Phi-4-Mini-Reasoning-Abliterated-GGUF: Quantization
- mradermacher/Phi-4-Mini-Reasoning-Abliterated-GGUF: Same weights, converted