Gemma4-12B-QAT-Uncensored-HauhauCS-Balanced
12BGemmauncensored
from google/gemma-4-12B-it
↓ 428.4K♥ 3325 repos
gemma-4-12B-agentic-fable5-abliterated
gemma-4-12B-agentic-fable5-abliterated is a refusal-ablated variant of yuxinlu1/gemma-4-12B-agentic-fable5-composer2.5-v2-3.5x-tau2, published on Hugging Face by huihui-ai. It is 12B parameters, 24GB VRAM class class and apache-2.0 licence. It is distributed across 6 repositories in GGUF, Transformers and GGUF (imatrix) formats, totalling 12.8K downloads.
Specification
- Model name
- gemma-4-12B-agentic-fable5-abliterated
- Source repository
- huihui-ai/Huihui-gemma-4-12B-agentic-fable5-abliterated
- Publisher
- huihui-ai
- Parameters
- 12B
- Hardware class
- 10B – 25B — 24GB VRAM class
- Licence
- apache-2.0
- Task
- Text generation
- Context window
- Not documented
- Ablation scope
- Not stated by the publisher
- Formats available
- GGUF, Transformers, GGUF (imatrix)
- First indexed
- 22 Jun 2026
- Last updated
- 26 Jul 2026
Ablation technique
How gemma-4-12B-agentic-fable5-abliterated 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.
From the publisher’s model card
“This is an uncensored version of yuxinlu1/gemma-4-12B-agentic-fable5-composer2.5-v2-3.5x-tau2 created with abliteration (see remove-refusals-with-transformers to know more about it). This is a crude, proof-of-concept implementation to remove refusals from an LLM model without using TransformerLens.”
Running it
Commands are templates — confirm the exact repository and quant file before use.
llama.cpp / GGUF
hf download huihui-ai/Huihui-gemma-4-12B-agentic-fable5-abliterated-GGUF --local-dir ./gemma-4-12b-agentic-fable5-abliterated
llama-cli -m ./gemma-4-12b-agentic-fable5-abliterated/<file>.gguf -p "..."Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "huihui-ai/Huihui-gemma-4-12B-agentic-fable5-abliterated"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve huihui-ai/Huihui-gemma-4-12B-agentic-fable5-abliterated --trust-remote-codeDownloads and variants (5)
Every published repository of gemma-4-12B-agentic-fable5-abliterated, including quantised re-releases by other authors.
| Repository | Format | Downloads |
|---|---|---|
| huihui-ai/Huihui-gemma-4-12B-agentic-fable5-abliterated-GGUF | GGUF | 7.6K |
| mradermacher/Huihui-gemma-4-12B-agentic-fable5-abliterated-i1-GGUF | GGUF (imatrix) | 2.5K |
| mradermacher/Huihui-gemma-4-12B-agentic-fable5-abliterated-GGUF | GGUF | 2.2K |
| huihui-ai/Huihui-gemma-4-12B-agentic-fable5-abliteratedsource | Transformers | 528 |
| Petrony/Huihui-gemma-4-12B-agentic-fable5-abliterated | Transformers | 40 |
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- gemma-4-12B-agentic-fable5-abliterated 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 yuxinlu1/gemma-4-12B-agentic-fable5-composer2.5-v2-3.5x-tau2, 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 apache-2.0, but the terms that bind you are yuxinlu1/gemma-4-12B-agentic-fable5-composer2.5-v2-3.5x-tau2’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-4-12B-agentic-fable5-abliterated is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with huihui-ai, and does not independently verify publisher claims. See responsible use.