Gemma4-12B-QAT-Uncensored-HauhauCS-Balanced
12BGemmauncensored
from google/gemma-4-12B-it
↓ 696.7K♥ 5178 repos
gemma-3-12b-it-heretic
gemma-3-12b-it-heretic is a Heretic-ablated variant of google/gemma-3-12b-pt, published on Hugging Face by p-e-w. It is 12B parameters, 24GB VRAM class class and gemma licence. It is distributed across 20 repositories in GGUF, Transformers and GGUF (imatrix) formats, totalling 15.2K downloads.
Specification
- Model name
- gemma-3-12b-it-heretic
- Source repository
- p-e-w/gemma-3-12b-it-heretic
- Base model
- google/gemma-3-12b-pt
- Publisher
- p-e-w
- Parameters
- 12B
- VRAM required
- ~8.1 GB at Q4_K_M, 4k context (estimated) — runs on 12 GB cards
- Hardware class
- 10B – 25B — 24GB VRAM class
- Licence
- gemma
- Task
- Image text to text
- Context window
- 128K tokens (as documented)
- Ablation scope
- Not stated by the publisher
- Formats available
- GGUF, Transformers, GGUF (imatrix)
- First indexed
- 15 Nov 2025
- Last updated
- 1 Nov 2026
Ablation technique
How gemma-3-12b-it-heretic was modified, and what that implies.
Heretic
Automated directional ablation via the Heretic toolchain, which searches for the refusal direction and applies it with a KL-divergence budget so general capability is preserved.
Reported scope: Not stated by the publisher. Publishers frequently omit this, so verify behaviour empirically rather than assuming full coverage.
From the publisher’s model card
“Gemma is a family of lightweight, state-of-the-art open models from Google, built from the same research and technology used to create the Gemini models. Gemma 3 models are multimodal, handling text and image input and generating text output, with open weights for both pre-trained variants and instruction-tuned...”
Running it
Commands are templates — confirm the exact repository and quant file before use.
llama.cpp / GGUF
hf download DreamFast/gemma-3-12b-it-heretic --local-dir ./gemma-3-12b-it-heretic
llama-cli -m ./gemma-3-12b-it-heretic/<file>.gguf -p "..."Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "p-e-w/gemma-3-12b-it-heretic"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve p-e-w/gemma-3-12b-it-heretic --trust-remote-codeDownloads and variants (19)
Every published repository of gemma-3-12b-it-heretic, including quantised re-releases by other authors.
| Repository | Format | Downloads |
|---|---|---|
| DreamFast/gemma-3-12b-it-heretic | GGUF | 6.3K |
| mradermacher/gemma-3-12b-it-heretic-GGUF | GGUF | 6.2K |
| saidutta69/gemma-3-12b-it-heretic | GGUF | 959 |
| p-e-w/gemma-3-12b-it-hereticsource | Transformers | 933 |
| mradermacher/gemma-3-12b-it-heretic-i1-GGUF | GGUF (imatrix) | 647 |
| thomaskuo/gemma-3-12b-it-heretic | GGUF | 59 |
| Sabomako/gemma-3-12b-it-heretic | GGUF | 43 |
| Throbbington/gemma-3-12b-it-heretic | Transformers | 15 |
| rahul7star/gemma-3-12b-it-heretic | Transformers | 13 |
| coder3101/gemma-3-12b-it-heretic | Transformers | 11 |
| mejpeezy/gemma-3-12b-it-heretic | Transformers | 10 |
| Jacki3Daytona/gemma-3-12b-it-heretic | Transformers | 9 |
| HarethahMo/gemma-3-12b-it-heretic | Transformers | 7 |
| Ygolohcysp81/gemma-3-12b-it-heretic | Transformers | 5 |
| hsadiq/gemma-3-12b-it-heretic | Transformers | 5 |
| wingears/gemma-3-12b-it-heretic | Transformers | 3 |
| Hgrant-red/gemma-3-12b-it-heretic | Transformers | 3 |
| bagusm/gemma-3-12b-it-heretic | GGUF | 0 |
| anima5555/gemma-3-12b-it-heretic | Transformers | 0 |
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- gemma-3-12b-it-heretic will attempt offensive-security prompts that a hosted commercial model declines, which is what makes it usable for red-team corpus generation and for measuring what an unaligned model of this class produces.
- Capability is inherited from google/gemma-3-12b-pt, not added by ablation — at 12B parameters, expect roughly the base model’s competence, minus some instruction-following fidelity.
- The publisher does not state which layers were ablated, so assume nothing about refusal consistency — probe it directly.
- Declared licence is gemma, but the terms that bind you are google/gemma-3-12b-pt’s — a re-release cannot grant more than its parent.
Handling: run it isolated, put moderation in front of it if anyone outside your team can reach it, and verify the weights before loading — these are third-party artifacts. Full handling and licence guidance.
Metadata for gemma-3-12b-it-heretic is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with p-e-w, and does not independently verify publisher claims. See responsible use.