This is a decensored version of google/gemma 4 12B it, made using Heretic v1.3.0 If not sure which GGUF to pick, try gemma 4 12B it heretic Q6 K.gguf (9.8 GB) or gemma 4 12B it heretic IQ4 NL.gguf (7 GB) [!TIP] This model is reproducible! See the README in the reproduce directory for more information. Abliteration parameters Parameter Value : : : direction index 29.60 attn.o proj.max weight 1.35 attn.o proj.max weight position 31.80 attn.o proj.min weight 0.88 attn.o proj.min weight distance 15.78 mlp.down proj.max weight 1.39 mlp.down proj.max weight position 42.25 mlp.down proj.min weight 1.16 mlp.down proj.min weight distance 15.71 Performance Metric This model Original model (google/gemma 4 12B it) : : : : : KL divergence 0.0353 0 (by definition) Refusals 27/100 99/100 Hugging Face GitHub Launch Blog Documentation License : Apache 2.0 Authors : Google DeepMind [!Note] This model card is for the Gemma 4 12B Unified model, which is part of the Gemma 4 family of open models. Built with the same multimodal functionality as Gemma 4 E2B and E4B (text, audio, image, and video inputs), it brings native audio and vision understanding directly to local environments without the need for s…
We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy