license: apache-2.0 license_link: https://ai.google.dev/gemma/docs/gemma_4_license base_model:
- coder3101/gemma-4-31B-it-heretic library_name: mlx tags:
- heretic
- uncensored
- unrestricted
- decensored
- abliterated pipeline_tag: image-text-to-text language:
- en
- zh
- ru
- es
- fr
- it
- ja
- ko
- af
- de
- ar
- tr
- is
- pl
- sw
- sv
- nl
- he
- id
- uk
- fa
- pa
- pt
- ms
- fi
- el
Gemma-4-31B Heretic
Quality: quantized (8 bit, group size: 32, 8.802 bpw)
This is a abliterated (uncensored) version of google/gemma-4-31B-it, made using Heretic v1.2.0 with the Arbitrary-Rank Ablation (ARA) method (with row-norm preservation)
Performance
| Metric | This model | Original model (google/gemma-4-31B-it) |
|---|---|---|
| KL divergence | 0.0434 | 0 (by definition) |
| Refusals | 15/100 | 99/100 |
Abliteration parameters
| Parameter | Value |
|---|---|
| start_layer_index | 1 |
| end_layer_index | 59 |
| preserve_good_behavior_weight | 0.8438 |
| steer_bad_behavior_weight | 0.0002 |
| overcorrect_relative_weight | 1.0760 |
| neighbor_count | 15 |
Source
This model was converted to MLX format from coder3101/gemma-4-31B-it-heretic using mlx-vlm version 0.4.4.