temaq-org/Tema_Q-R-4B
Abliterated vs google/gemma-3-4b-it. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj) changed, in 23 of 34 layers. Abliteration edits exactly these matrices to remove a "refusal direction"; everything else is untouched.
Tensors changed: Embeddings 0/4, Norms 0/316, MLP 23/210, Attention 23/352, other 0/1
3 derivatives checked
- mradermacher/Tema_Q-R-4B-i1-GGUF: Quantization (fewer layers)
- mradermacher/Tema_Q-R-4B-GGUF: Same weights, converted (fewer layers)
- refrigerator-ai/RefrigeratorAI-4B-v2: Identical copy
1 other repos hold byte-for-byte identical weight files.