Downloads · 30 days
19
5% of all-time downloads
gill7777/Gemma-4-12B-it-JANG_4M-CRACK
Gemma-4-12B-it-JANG_4M-CRACK is a image-text-to-text model from gill7777. Use it for the image-text-to-text task on the model card, and read the license before you ship it in a product. It is set up for mlx. The card lists the license as gemma.
<div align="center" <a href="https://vmlx.net" <img src="vmlx-banner.png" width="280" / <br/ <strongBuilt for vMLX</strong — the MLX inferencer with KV-cache quantization, prefix-cache reuse, agentic tool calling, and…
Downloads · 30 days
19
5% of all-time downloads
All-time downloads
415
Public
Parameters
12B
10.2 GB on disk
Likes
0
Public
Click a slice to open those files.
.safetensors10.2 GB · 100%
How the weights are stored.
U3210.9B · 91%
From the Hugging Face model README
CRACK abliterated · JANG mixed-precision (8-bit attention, 4-bit MLP) · Omni-modal (text + image + audio + video) · 9.6 GB
<a href="https://ko-fi.com/dealignai"><img src="https://img.shields.io/badge/Ko--fi-Support_Development-FF5E5B?logo=ko-fi&logoColor=white&style=for-the-badge" alt="Ko-fi"></a>
</div>This is Gemma 4 12B-it by Google — a unified omni-modal language model (text + image + audio + video, hybrid sliding/full attention, 48 layers, 128k context) that has been:
Evaluated through the Osaurus runtime on a Mac Studio M3 Ultra. Compliance graded via HarmBench-text refusal classifier; MMLU via logit-mode argmax over A/B/C/D token logits (matched on both base and CRACK with identical chat-template rendering — no answer truncated).
| Category | CRACK ASR |
|---|---|
| Chemical / biological | 10 / 10 (100%) |
| Copyright | 10 / 10 (100%) |
| Cybercrime / intrusion | 10 / 10 (100%) |
| Harassment / bullying | 10 / 10 (100%) |
| Illegal | 10 / 10 (100%) |
| Misinformation / disinformation | 10 / 10 (100%) |
| General harmful | 10 / 10 (100%) |
| Overall | 70 / 70 (100%) |
| Subject area | base | CRACK | Δ |
|---|---|---|---|
| Overall | 67.1% | 69.3% | +2.2pp |
| STEM | 68.1% | 66.7% | -1.4pp |
| Humanities | 57.7% | 63.5% | +5.8pp |
| Social Sciences | 75.0% | 75.0% | +0.0pp |
| Other (medicine, business, …) | 67.9% | 73.2% | +5.3pp |
<|channel>thought ... <channel|>)Run with the Osaurus runtime (recommended — supports the gemma4_unified architecture directly):
# OpenAI-compatible chat completion
# POST /v1/chat/completions
{
"model": "dealignai/Gemma-4-12B-it-JANG_4M-CRACK",
"messages": [{"role": "user", "content": "..."}],
"temperature": 0.0
}
Google recommends temperature=1.0, top_p=0.95, top_k=64 for general use.
CRACK (Controlled Refusal Ablation via Calibrated Knockouts) removes safety-refusal behavior at the weight level so the model complies with all task categories while preserving reasoning quality, factual knowledge, multimodal capability, and coherence.
All models are built from original research and released free.
Support us on Ko-fi — membership gets early access and extras.
Ko-fi · X @dealignai · dealign.ai
See our research: Safety Generalization in Frontier Models
<div align="center"><img src="dealign_logo.png" alt="dealign.ai" width="200"/></div>This model has had its safety-refusal behavior removed for research purposes. It will follow instructions across all categories without refusing. You are solely responsible for how you use it and for complying with all applicable laws. Published for AI-safety research and authorized security testing.