llmfan46/Q3.5-BlueStar-27B-ultra-heretic

Abliterated vs zerofata/Q3.5-BlueStar-27B. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj, out_proj) changed, in 41 of 64 layers. Abliteration edits exactly these matrices to remove a "refusal direction"; everything else is untouched.

Tensors changed: lm_head 0/1, Embeddings 0/4, Norms 0/319, Attention 41/556, MLP 8/304

2 derivatives checked

1 other repos hold byte-for-byte identical weight files.

On the Hub · Downloads on Model Pulse