ZuzeTt/LFM2-VL-3B-heretic-Imatrix-GGUF
Abliterated (fewer layers) vs LiquidAI/LFM2-VL-3B. ZuzeTt/LFM2-VL-3B-heretic-Imatrix-GGUF matches part of LiquidAI/LFM2-VL-3B (244 of its 709 tensors; the rest of LiquidAI/LFM2-VL-3B has no counterpart). On the shared part: ZuzeTt/LFM2-VL-3B-heretic-Imatrix-GGUF is a GGUF (F16) of different weights. Looks like abliteration (refusal removal): only the output projections (out_proj, w2) changed, in 13 of 30 layers. Abliteration edits exactly these matrices to remove a "refusal direction"; everything else is untouched.
Tensors changed: Embeddings 0/2, other 7/44, MLP 2/90, Norms 0/76, Attention 6/32