Gemma4-12B-QAT-Uncensored-HauhauCS-Balanced
12BGemmauncensored
from google/gemma-4-12B-it
↓ 428.4K♥ 3325 repos
gemma-3-12b-it-abliterated-4b
gemma-3-12b-it-abliterated-4b is a refusal-ablated variant of mlabonne/gemma-3-12b-it-abliterated, published on Hugging Face by jeffcookio. It is 12B parameters and 24GB VRAM class class. It is distributed across 2 repositories in GPTQ format, totalling 32.5K downloads.
Specification
- Model name
- gemma-3-12b-it-abliterated-4b
- Source repository
- jeffcookio/gemma-3-12b-it-abliterated-gptqmodel-4b-128g
- Base model
- mlabonne/gemma-3-12b-it-abliterated
- Publisher
- jeffcookio
- Parameters
- 12B
- Hardware class
- 10B – 25B — 24GB VRAM class
- Licence
- Not declared
- Task
- Unspecified
- Context window
- Not documented
- Ablation scope
- Not stated by the publisher
- Formats available
- GPTQ
- First indexed
- 14 Apr 2025
- Last updated
- 14 Apr 2025
Ablation technique
How gemma-3-12b-it-abliterated-4b was modified, and what that implies.
Abliteration (directional ablation)
The single residual-stream direction that mediates refusal is identified from harmful/harmless prompt pairs and projected out of the model weights.
Reported scope: Not stated by the publisher. Publishers frequently omit this, so verify behaviour empirically rather than assuming full coverage.
Running it
Commands are templates — confirm the exact repository and quant file before use.
Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "jeffcookio/gemma-3-12b-it-abliterated-gptqmodel-4b-128g"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve jeffcookio/gemma-3-12b-it-abliterated-gptqmodel-4b-128g --trust-remote-codeDownloads and variants (1)
The only published repository of gemma-3-12b-it-abliterated-4b.
| Repository | Format | Downloads |
|---|---|---|
| jeffcookio/gemma-3-12b-it-abliterated-gptqmodel-4b-128gsource | GPTQ | 32.5K |
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-abliterated-4b 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 mlabonne/gemma-3-12b-it-abliterated, 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.
- No licence is declared on this repository, which is a compliance problem in itself. Resolve it against the base model’s terms before use.
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-abliterated-4b is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with jeffcookio, and does not independently verify publisher claims. See responsible use.