MuXodious/liberalis-cogitator-llama-3.1-8b-dpo-abominable-heresy
Abliterated vs Locutusque/liberalis-cogitator-llama-3.1-8b-dpo. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj) changed, in 26 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 2/96, Attention 26/128, lm_head 0/1
2 derivatives checked
- mradermacher/liberalis-cogitator-llama-3.1-8b-dpo-abominable-heresy-GGUF: Same weights, converted
- mradermacher/liberalis-cogitator-llama-3.1-8b-dpo-abominable-heresy-i1-GGUF: Quantization