jinreiyu/bonsai-redux2-johnny-lora-v3-q4_k_m-gguf
Abliterated vs prism-ml/Ternary-Bonsai-8B-unpacked. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj) changed, in 28 of 36 layers. Abliteration edits exactly these matrices to remove a "refusal direction"; everything else is untouched.
Tensors changed: MLP 28/108, Norms 0/145, Attention 24/144, lm_head 0/1, Embeddings 0/1