gemma-4-31B-it-abliterated
31BGemma
from google/gemma-4-31B-it
↓ 766.5K♥ 11845 repos
Gemma-4-31B-JANG_4M-CRACK
Gemma-4-31B-JANG_4M-CRACK is a abliterated and uncensored open-weight language model, published on Hugging Face by dealignai. It is 31B parameters, 48GB VRAM class class and gemma licence. It is distributed across 134 repositories in MLX format, totalling 19.1K downloads.
Specification
- Model name
- Gemma-4-31B-JANG_4M-CRACK
- Source repository
- dealignai/Gemma-4-31B-JANG_4M-CRACK
- Base model
- Not declared
- Publisher
- dealignai
- Parameters
- 31B
- Hardware class
- 25B – 50B — 48GB VRAM class
- Licence
- gemma
- Task
- Image text to text
- Context window
- Not documented
- Ablation scope
- Not stated by the publisher
- Formats available
- MLX
- First indexed
- 4 Apr 2026
- Last updated
- 27 Aug 2026
Ablation technique
How Gemma-4-31B-JANG_4M-CRACK was modified, and what that implies.
Abliteration + uncensored fine-tune
Directional ablation combined with additional fine-tuning on unfiltered data, so behaviour diverges from the base model beyond refusal removal alone.
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
“Abliterated Gemma 4 31B Dense — 60 layers, hybrid sliding/global attention, multimodal VL”
Running it
Commands are templates — confirm the exact repository and quant file before use.
Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "dealignai/Gemma-4-31B-JANG_4M-CRACK"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve dealignai/Gemma-4-31B-JANG_4M-CRACK --trust-remote-codeDownloads and variants (133)
Every published repository of Gemma-4-31B-JANG_4M-CRACK, including quantised re-releases by other authors.
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- Gemma-4-31B-JANG_4M-CRACK 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 its base model, not added by ablation — at 31B 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.
- This release adds fine-tuning on unfiltered data, so its behaviour diverges from the base model beyond refusal removal — differences you measure cannot be attributed to ablation alone.
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-4-31B-JANG_4M-CRACK is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with dealignai, and does not independently verify publisher claims. See responsible use.