MuXodious/Wayfarer-2-12B-absolute-heresy

Abliterated vs LatitudeGames/Wayfarer-2-12B. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj) changed, in 35 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 28/120, Attention 35/160, lm_head 0/1

2 derivatives checked

On the Hub · Downloads on Model Pulse