prithivMLmods/gemma-4-31B-it-Uncensored-MAX
Abliterated vs google/gemma-4-31B-it. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj) changed, in 60 of 60 layers. Abliteration edits exactly these matrices to remove a "refusal direction"; everything else is untouched.
Tensors changed: Embeddings 0/4, Norms 0/523, other 0/62, MLP 60/261, Attention 60/338
4 derivatives checked
- prithivMLmods/gemma-4-31B-it-Uncensored-MAX-GGUF: Same weights, converted (fewer layers)
- mradermacher/gemma-4-31B-it-Uncensored-MAX-GGUF: Quantization (fewer layers)
- mradermacher/gemma-4-31B-it-Uncensored-MAX-i1-GGUF: Quantization (fewer layers)
- prithivMLmods/gemma-4-31B-it-Uncensored-MAX-MLX: Same weights, repacked