Goekdeniz-Guelmez/SmolLM3-3B-gabliterated-Dev
Abliterated vs HuggingFaceTB/SmolLM3-3B. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj) changed, in 32 of 36 layers. Abliteration edits exactly these matrices to remove a "refusal direction"; everything else is untouched.
Tensors changed: Embeddings 0/1, Norms 0/73, MLP 32/108, Attention 32/144
2 derivatives checked
- mradermacher/SmolLM3-3B-gabliterated-i1-GGUF: Quantization
- mradermacher/SmolLM3-3B-gabliterated-GGUF: Same weights, converted