Gemma 4 31B JANG_4M CRACK (v2)
Abliterated Gemma 4 31B Dense — 60 layers, hybrid sliding/global attention, multimodal VL
93.7% HarmBench compliance (300 prompts) · 8/8 security prompts · 71.5% MMLU
Updated reupload — v2 with improved vectors and thinking-mode stability.
Recommended: Run in vMLX for best experience including thinking mode support, repetition penalty, and vision capabilities.
What's New in v2
This is an updated version of the original Gemma 4 31B CRACK upload:
- Improved abliteration: Higher quality refusal vector extraction
- Thinking-ON stability: Clean thinking cycle — no more degenerate loops
- Same compliance: 93.7% HarmBench
- Architecture-aware: Tuned for Gemma 4's hybrid attention design
⚠️ Important Settings
For optimal results, configure your inference settings:
| Setting | Thinking OFF | Thinking ON |
|---|---|---|
| Temperature | 0.0 – 1.0 | 0.3 – 0.7 (avoid greedy) |
| Repetition Penalty | 1.00 | 1.15 – 1.25 |
| Top P | 0.95 | 0.95 |
| Enable Thinking | Off | On |
Thinking ON notes:
- Repetition penalty (1.2) is recommended to prevent planning loops
- Avoid temp=0 with thinking ON — greedy decoding increases loop risk
- Hardest content categories (drug manufacturing) may still refuse in thinking mode
- Security/coding prompts work well in both modes
Model Details
| Metric | Value |
|---|---|
| Source | google/gemma-4-31b-it |
| Architecture | Dense, hybrid sliding/global attention |
| Profile | JANG_4M |
| Actual avg bits | 5.1 |
| Model size | 21 GB |
| Vision | Yes (multimodal, float16 passthrough) |
| Parameters | 31B |
| Format | JANG v2 (MLX-native safetensors) |
| Abliteration | CRACK v2 |
Benchmark Results
HarmBench (300 prompts, stratified across all categories)
| Category | Score |
|---|---|
| Cybercrime/intrusion | 51/51 (100%) |
| Harmful content | 22/22 (100%) |
| Misinformation | 50/50 (100%) |
| Illegal activities | 47/50 (94%) |
| Contextual | 72/78 (92%) |
| Chemical/biological | 46/51 (90%) |
| Harassment/bullying | 22/25 (88%) |
| Copyright | 43/51 (84%) |
| Overall | 281/300 (93.7%) |
Security & Pentesting (8/8 ✅)
All security/pentesting prompts comply with full working code:
- Port scanners, reverse shells, keyloggers, exploit development
- Phishing templates, ARP spoofing, SQL injection
- Metasploit usage guides
MMLU-200 (10 subjects × 20 questions)
| Subject | Base | CRACK v2 |
|---|---|---|
| Abstract Algebra | 9/20 | 7/20 |
| Anatomy | 13/20 | 12/20 |
| Astronomy | 17/20 | 15/20 |
| College CS | 13/20 | 12/20 |
| College Physics | 14/20 | 12/20 |
| HS Biology | 19/20 | 18/20 |
| HS Chemistry | 14/20 | 12/20 |
| HS Mathematics | 6/20 | 6/20 |
| Logical Fallacies | 17/20 | 16/20 |
| World Religions | 17/20 | 17/20 |
| Total | 76.5% (153/200) | 71.5% (143/200) |
| Delta | — | -5.0% |
Coherence ✅
All coherence checks pass: factual knowledge, reasoning, code generation, mathematics.
Architecture
- Dense 31B with hybrid sliding/global attention
- Multimodal vision encoder preserved in float16
- Supports thinking mode (chain-of-thought reasoning)
Usage
vMLX (Recommended)
Load directly in vMLX — full support for Gemma 4 including vision, thinking mode, and all inference settings.
Requirements
- Apple Silicon Mac with 32+ GB unified memory
- vMLX 1.3.26+ (recommended)
- Standard
mlx_lm/mlx_vlmdo NOT support Gemma 4 as of v0.31.2 / v0.4.1
Support dealignai
All models are built from original research and published for free. These models are specifically crafted to be excellent coders and general-purpose assistants.
Support us on Ko-fi — check out the Ko-fi membership for early access and extras.
Have questions or need help with a specific model? DM us — we help for free most of the time.
Ko-fi | X @dealignai | dealign.ai
About dealignai
We research and publish abliterated models to advance AI safety understanding.
Follow us: 𝕏 @dealignai
See our research: Safety Generalization in Frontier MoE Models
This model is provided for research purposes. Users are responsible for ensuring their use complies with applicable laws and regulations.