-
gemma-4-12B-it-heretic-GGUF
Fully decensored Gemma 4 12B (abliterated with Heretic) — 0/100 genuine refusals at KL 0.0284, i.e. near-zero capability loss.
tools thinking19.7K Pulls 3 Tags Updated 4 months ago
-
gemma-4-E4B-it-heretic-GGUF
Abliterated (decensored) version of google/gemma-4-E4B-it, created using Heretic v1.2.0 with Arbitrary-Rank Ablation (ARA) and row-norm preservation.
tools thinking14.5K Pulls 5 Tags Updated 6 months ago
-
gemma-4-12B-it-qat-q4_0-unquantized-heretic
Decensored (abliterated) Gemma 4 12B — refusals removed with Heretic while keeping the base model's intelligence (KL 0.0154, 0 true refusals). For uncensored assistant and 18+ companion/roleplay use.
tools thinking10.3K Pulls 4 Tags Updated 4 months ago
-
gemma-4-E4B-it-qat-q4_0-unquantized-heretictools thinking
1,293 Pulls 3 Tags Updated 4 months ago
-
gemma4-e4b-classifier
A modality-stripped variant of google/gemma-4-E4B-it for text-only classification, entity extraction, and structured-memory extraction. The vision encoder (150M params) and audio encoder (300M params) are removed; the text path is unchanged.
tools thinking367 Pulls 3 Tags Updated 5 months ago