mradermacher/gemma-4-E4B-it-SOMPOA-heresy-GGUF
Abliterated (fewer layers) vs MuXodious/gemma-4-E4B-it-SOMPOA-heresy. mradermacher/gemma-4-E4B-it-SOMPOA-heresy-GGUF matches part of MuXodious/gemma-4-E4B-it-SOMPOA-heresy (623 of its 2130 tensors; the rest of MuXodious/gemma-4-E4B-it-SOMPOA-heresy has no counterpart). On the shared part: mradermacher/gemma-4-E4B-it-SOMPOA-heresy-GGUF is a GGUF (F16) of different weights. Looks like abliteration (refusal removal): only the output projections (down_proj, o_proj) changed, in 19 of 42 layers. Abliteration edits exactly these matrices to remove a "refusal direction"; everything else is untouched.
Tensors changed: Embeddings 0/2, Norms 0/278, MLP 8/126, other 0/85, Attention 16/132