jwest33/gemma-3-4b-it-null-space-abliterated
Full fine-tune vs google/gemma-3-4b-it. Fine-tune: 90% of the weights changed, across 34 of 34 layers.
Tensors changed: Embeddings 1/4, Norms 0/316, MLP 102/210, Attention 135/352, other 0/1
3 derivatives checked
- jwest33/gemma-3-4b-null-space-abliterated-RP-writer-GGUF: LoRA-style (fewer layers)
- mradermacher/gemma-3-4b-it-null-space-abliterated-i1-GGUF: Quantization (fewer layers)
- mradermacher/gemma-3-4b-it-null-space-abliterated-GGUF: Full fine-tune (fewer layers)