grayarea/Magidonia-24B-v4.3-heretic-v1.2
Abliterated vs TheDrummer/Magidonia-24B-v4.3. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj) changed, in 34 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 34/120, Attention 29/160, lm_head 0/1
3 derivatives checked
- mradermacher/Magidonia-24B-v4.3-heretic-v1.2-i1-GGUF: Quantization
- mradermacher/Magidonia-24B-v4.3-heretic-v1.2-GGUF: Quantization
- grayarea/Magidonia-24B-v4.3-heretic-v1.2-GGUF: Same weights, converted