DuoNeural/Mistral-NeMo-12B-Abliterated
Abliterated vs mistralai/Mistral-Nemo-Instruct-2407. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj) changed, in 40 of 40 layers. Abliteration edits exactly these matrices to remove a "refusal direction"; everything else is untouched.
Tensors changed: Embeddings 0/1, Norms 0/81, MLP 40/120, Attention 40/160, lm_head 0/1
2 derivatives checked
- mradermacher/Mistral-NeMo-12B-Abliterated-i1-GGUF: Quantization
- mradermacher/Mistral-NeMo-12B-Abliterated-GGUF: Quantization