treadon/granite-4.1-3b-Abliterated-AND-Disinhibited
Abliterated vs ibm-granite/granite-4.1-3b. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj) changed, in 20 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 20/120, Attention 20/160
3 derivatives checked
- mradermacher/granite-4.1-3b-Abliterated-AND-Disinhibited-i1-GGUF: Quantization
- Dzluck/granite-4.1-3b-Abliterated-AND-Disinhibited-i1-GGUF: Quantization
- mradermacher/granite-4.1-3b-Abliterated-AND-Disinhibited-GGUF: Same weights, converted