richardyoung/zephyr-7b-beta-abliterated
Abliterated vs HuggingFaceH4/zephyr-7b-beta. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj) changed, in 19 of 32 layers. Abliteration edits exactly these matrices to remove a "refusal direction"; everything else is untouched.
Tensors changed: Embeddings 0/1, Norms 0/65, MLP 18/96, Attention 9/128, lm_head 0/1
2 derivatives checked
- mradermacher/zephyr-7b-beta-abliterated-i1-GGUF: Quantization
- mradermacher/zephyr-7b-beta-abliterated-GGUF: Same weights, converted