Gemma-4-E4B-Uncensored-HauhauCS-Aggressive
Gemmauncensored
from HauhauCS/Gemma-4-E4B-Uncensored-HauhauCS-Aggressive
↓ 757.6K♥ 1.5K51 repos
gemma4-E2B-it-qat-unsloth-heretic
gemma4-E2B-it-qat-unsloth-heretic is a Heretic-ablated variant of unsloth/gemma-4-E2B-it-qat-q4_0-unquantized, published on Hugging Face by MobiusDevelopment. It is apache-2.0 licence. It is distributed across 2 repositories in GGUF format, totalling 5.4K downloads.
Specification
- Model name
- gemma4-E2B-it-qat-unsloth-heretic
- Source repository
- MobiusDevelopment/gemma4-E2B-it-qat-Q4-unsloth-heretic
- Publisher
- MobiusDevelopment
- Parameters
- Not determinable from name
- VRAM required
- Unknown — no published parameter count
- Hardware class
- Unknown
- Licence
- apache-2.0
- Task
- Text generation
- Context window
- Not documented
- Ablation scope
- Not stated by the publisher
- Formats available
- GGUF
- First indexed
- 23 Jun 2026
- Last updated
- 23 Jun 2026
Ablation technique
How gemma4-E2B-it-qat-unsloth-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
“This is a decensored ("abliterated") build of Google's Gemma 4 E2B instruction-tuned model. The refusal behavior has been removed so the model answers prompts it would otherwise decline, while its general capabilities are kept intact.”
Running it
Commands are templates — confirm the exact repository and quant file before use.
Ollama (publisher-documented)
ollama run MobiusDevelopment/gemma4-E2B-it-qat-Q4-unsloth-hereticllama.cpp / GGUF
hf download MobiusDevelopment/gemma4-E2B-it-qat-Q4-unsloth-heretic --local-dir ./gemma4-e2b-it-qat-unsloth-heretic
llama-cli -m ./gemma4-e2b-it-qat-unsloth-heretic/<file>.gguf -p "..."Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "MobiusDevelopment/gemma4-E2B-it-qat-Q4-unsloth-heretic"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve MobiusDevelopment/gemma4-E2B-it-qat-Q4-unsloth-heretic --trust-remote-codeDownloads and variants (1)
The only published repository of gemma4-E2B-it-qat-unsloth-heretic.
| Repository | Format | Downloads |
|---|---|---|
| MobiusDevelopment/gemma4-E2B-it-qat-Q4-unsloth-hereticsource | GGUF | 5.4K |
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- gemma4-E2B-it-qat-unsloth-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.
- 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 unsloth/gemma-4-E2B-it-qat-q4_0-unquantized’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 gemma4-E2B-it-qat-unsloth-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 MobiusDevelopment, and does not independently verify publisher claims. See responsible use.