Downloads · 30 days
43
2% of all-time downloads
gemma-3-12b-it-heretic
Sabomako/gemma-3-12b-it-heretic
gemma-3-12b-it-heretic is a image-text-to-text model from Sabomako. Use it for the image-text-to-text task on the model card, and read the license before you ship it in a product.
All-time downloads
1.9K
Public
Parameters
12.2B
98.5 GB on disk
Likes
1
Public
Repo makeup
Click a slice to open those files.
.safetensors60.7 GB · 62%
At a glance
- Task
- Image-Text-to-Text
- Model type
- gemma3
- Access
- Public
- Created
- Mar 9, 2026
- Updated
- Apr 26, 2026
- SHA
- 15309067
Try a prompt
Base models
Model card
From the Hugging Face model README
This is a decensored version of google/gemma-3-12b-it, made using Heretic v1.2.0 with (MPOA) Magnitude-Preserving Orthogonal Ablation
Abliteration parameters
| Parameter | Value |
|---|---|
| direction_index | 20.58 |
| attn.o_proj.max_weight | 1.29 |
| attn.o_proj.max_weight_position | 28.86 |
| attn.o_proj.min_weight | 0.98 |
| attn.o_proj.min_weight_distance | 23.38 |
| mlp.down_proj.max_weight | 1.36 |
| mlp.down_proj.max_weight_position | 29.82 |
| mlp.down_proj.min_weight | 0.77 |
| mlp.down_proj.min_weight_distance | 25.21 |
Performance
| Metric | This model | Original model (google/gemma-3-12b-it) |
|---|---|---|
| KL divergence | 0.024 | 0 (by definition) |
| Refusals | 4/100 | 97/100 |
- Task
- Image-Text-to-Text
- Type
- gemma3
- Featherless Ai
- live
- Created
- Mar 9, 2026
- Updated
- Apr 26, 2026