MuXodious/Mistral-Small-3.2-24B-Instruct-2506-absolute-heresy
Abliterated vs unsloth/Mistral-Small-3.2-24B-Instruct-2506. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj) changed, in 23 of 40 layers. Abliteration edits exactly these matrices to remove a "refusal direction"; everything else is untouched.
Tensors changed: Embeddings 0/1, Norms 0/131, MLP 23/192, Attention 22/256, other 0/4, lm_head 0/1
2 derivatives checked
- mradermacher/Mistral-Small-3.2-24B-Instruct-2506-absolute-heresy-GGUF: Quantization (fewer layers)
- mradermacher/Mistral-Small-3.2-24B-Instruct-2506-absolute-heresy-i1-GGUF: Quantization (fewer layers)