Get the app
← back to catalog · registered 2026-08-22 13:56

arnomatic/gemma-3-12b-it-4-k-m-gguf-heretic-abliterated

arnomatic Gemma 12B GGUF
Your rig guess connected
? Why do I need an app?
Reading your rig…

This is a rough estimate. Install the free app - we'll show exact numbers.

Reading real hardware from your app right now. Numbers below are exact.

Below is the per-quantization compatibility for this model. 

curl -H "Authorization: Bearer $ABL_KEY" \
     "https://abliteration.org/api/v1/models/arnomatic%2Fgemma-3-12b-it-4-k-m-gguf-heretic-abliterated"
Response includes
  • classification m3
  • files 4
  • author_summary 4 models
  • readme_text full
Try next
GET/models?author=arnomatic
All 4 models by arnomatic
 GET/models?family=gemma&method=m3
Sibling gemma models · same method
 GET/methods/m3
What "m3" means · taxonomy
10 credits · hourly refresh · ~4 KB payload Get an API key →
Abliteration classifier · v1.0.0
M3
Primary method

Layer-wise ablation

Applied on top of direct removal inherited from the base model.
Confidence
HIGH
Inherited from base model
M1 Direct removal →
Why this label 2 signals
Producer identity confirmed by naming conventions, tags or the model card. This label is very unlikely to change.
  • 'heretic' in model name (Heretic-produced)
  • Heretic uses layer-wise optimization (M3) with underlying direction removal (M1)
How we label → About M3 →
Refusal direction extracted via
Extraction technique

Difference-of-means

FromArditi et al. (2024) ↗
Confidence
HIGH
Why we say so
name contains 'heretic'; Heretic default extraction is difference-of-means (Arditi 2024)
What is a refusal direction? → Read the paper ↗
Downloads · 30-day
44
↑ 784% in 90 days
Likes
0
Model age
10mo ago
created 2025-11-23
Downloads over time
Now707→from80↑784%
4928952977080 on Nov 26, 2025707 on Sep 26Nov '25JanMarMayJulSep
Nov 26, 2025 → Sep 26 · 70 snapshots · spans 304 days

Metadata

License
gemma
Quantizations
Q4_K
Tags
gguf license:gemma endpoints_compatible region:us conversational

Related

Other by arnomatic
llama-3.3-70b-instruct-heretic-abliterated-Q4_K_M 110 gemma-3-12b-it-fp16-gguf-heretic-abliterated 47 gemma-3-27b-it-4-k-m-gguf-heretic-abliterated 44
Total size
6.80 GB
Files
4
Quantizations
3
Registered
2026-08-22 13:56
Last updated on HF
2025-11-23 15:29
Source huggingface.co/arnomatic/gemma-3-12b-it-4-k-m-gguf-heretic-abliterated ↗

Files by quantization

Q4_K 1 file 6.80 GB
gemma-3-12b-it-abliterated-text-Q4_K_M.gguf 6.80 GB a59a5a94 download
F16 1 file 815 MB
mmproj-gemma-3-12b-it-abliterated-fp16.gguf 815 MB b6a2899f download
Auxiliary files 2 files 1.67 KB
.gitattributes 1.64 KB c793919b download
README.md 26.0 B d6cac6e5 download

README current version from Hugging Face

license: gemma

README history 1 version

The author's README evolved over time. Click a version to see its content at that point.

  1. 2025-11-23initial commita3a904923 B
    Loading...
Catalog is the map. Apps are the tools.

Run models on your own machine, not in the cloud.

Every model page has an "Open in Abliteration" button that hands the model directly to the first-party desktop client, at the quantization your rig can actually run. No API keys, no subscription, no prompt leakage.

Get the app Browse the catalog
Open in Abliteration