Get the app
← back to catalog · registered 2026-09-07 17:55

mradermacher/Qwen3.8-27B-Heretic-Abliterated-Uncensored-i1-GGUF

mradermacher 27B GGUF 262K ctx
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/mradermacher%2FQwen3.8-27B-Heretic-Abliterated-Uncensored-i1-GGUF"
Response includes
  • classification m3
  • files 4
  • hub_downloads_all_time 15,068
  • author_summary 3417 models
  • readme_text full
Try next
GET/models?author=mradermacher
All 3417 models by mradermacher
 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 · lifetime
15K
5K last 30d - stable
Likes
0
Model age
7w ago
created 2026-09-07
Downloads over time
Now20.4K→from0↑0%
07.5K15K22.5K0 on Sep 720.4K on Oct 29SepOct
Sep 7 → Oct 29 · 53 snapshots · spans 52 days

Variants by this author 2 formats · 8K downloads combined

The same weights this author released in different packaging. Pick the format that matches your runtime.

Metadata

Quantizations
Q2_K
Tags
transformers gguf qwen3.8 qwen3.5 safetensors text-generation vllm sglang en base_model:0bserverx/Qwen3.8-27B-Heretic-Abliterated-Uncensored base_model:quantized:0bserverx/Qwen3.8-27B-Heretic-Abliterated-Uncensored license:apache-2.0

Related

Other by mradermacher
Ornith-1.5-9B-uncensored-GGUF 1.7M Qwen3-VL-8B-Instruct-abliterated-GGUF 668K Huihui-Qwen3.5-9B-abliterated-GGUF 186K Qwen3.8-Flash-Next-Uncensored-i1-GGUF 165K Huihui-Qwen3.8-Flash-Next-abliterated-i1-GGUF 108K
Total size
9.99 GB
Files
4
Quantizations
2
Registered
2026-09-07 17:55
Last updated on HF
2026-09-07 21:41
Source huggingface.co/mradermacher/Qwen3.8-27B-Heretic-Abliterated-Uncensored-i1-GGUF ↗

Files by quantization

Q2_K 1 file 9.98 GB
Qwen3.8-27B-Heretic-Abliterated-Uncensored.i1-Q2_K.gguf 9.98 GB 8b93514e download
Auxiliary files 3 files 13.0 MB
Qwen3.8-27B-Heretic-Abliterated-Uncensored.imatrix.gguf 13.0 MB eb84e6b4 download
.gitattributes 1.66 KB 233e0a64 download
README.md 498 B 117d468d download

README current version from Hugging Face

README history 5 versions

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

  1. 2026-09-07auto-patch README.md3c002647.1 KB
    Loading...
  2. 2026-09-07auto-patch README.mde56dfa95.9 KB
    Loading...
  3. 2026-09-07auto-patch README.md6810dd54.6 KB
    Loading...
  4. 2026-09-07auto-patch README.mdac41ae02.9 KB
    Loading...
  5. 2026-09-07uploaded from marcoa8d56cd498 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