huihui-ai/Huihui-MiniCPM5-2B-abliterated
Abliterated vs openbmb/MiniCPM5-2B. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj) changed, in 32 of 42 layers. Abliteration edits exactly these matrices to remove a "refusal direction"; everything else is untouched.
Tensors changed: lm_head 0/1, Embeddings 0/1, Norms 0/85, MLP 32/126, Attention 32/168
3 derivatives checked
- tinyopsec/Huihui-MiniCPM5-2B-abliterated-GGUF: Same weights, converted
- mradermacher/Huihui-MiniCPM5-2B-abliterated-i1-GGUF: Quantization
- mradermacher/Huihui-MiniCPM5-2B-abliterated-GGUF: Same weights, converted