Jiunsong/SuperGemma-4-12b-abliterated
LoRA-style vs google/gemma-4-12B-it. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, gate_proj, k_proj, o_proj, q_proj, up_proj, v_proj changed, in 48 of 48 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/12, Norms 0/289, other 0/48, MLP 144/144, Attention 184/184
8 derivatives checked
- mradermacher/SuperGemma-4-12b-abliterated-i1-GGUF: Quantization
- mradermacher/SuperGemma-4-12b-abliterated-GGUF: Quantization
- Jiunsong/SuperGemma-4-12b-abliterated-gguf-4bit: Quantization
- DreamFoundries/SuperGemma-4-12b-abliterated-4bit: Quantization
- Jiunsong/SuperGemma-4-12b-abliterated-mlx-4bit: Quantization
- DreamFoundries/SuperGemma-4-12b-abliterated-6bit: Quantization
- Jiunsong/SuperGemma-4-12b-abliterated-nvf4: Quantization
- DreamFoundries/SuperGemma-4-12b-abliterated-8bit: Quantization
1 other repos hold byte-for-byte identical weight files.