Get the app
← back to catalog · registered 2026-10-12 01:59

jirka642/Qwen3.8-27B-Uncensored-Heretic-Native-MTP-Preserved-exl3

jirka642 27B second-order
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/jirka642%2FQwen3.8-27B-Uncensored-Heretic-Native-MTP-Preserved-exl3"
Response includes
  • classification m3
  • files 2
  • author_summary 1 models
  • readme_text full
Try next
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
0
Likes
0
Model age
today
created 2026-10-12

Genealogy 0 direct forks

Full fork graph →

This model's place in the market. Above: what it was derived from. Below: the tree of everything derived from it.

llmfan46/Qwen3.8-27B-Uncensored-Heretic-Native-MTP-Preserved↓ 1,194 this lineage
jirka642/Qwen3.8-27B-Uncensored-Heretic-Native-MTP-Preserved-exl3

Metadata

Tags
exllamav3 safetensors text-generation base_model:llmfan46/Qwen3.8-27B-Uncensored-Heretic-Native-MTP-Preserved base_model:quantized:llmfan46/Qwen3.8-27B-Uncensored-Heretic-Native-MTP-Preserved region:us

Related

Same base model
codavidgarcia/Qwen3.8-27B-Uncensored-Heretic-MTP-UD-GGUF 2K sjoe1244/Qwen3.8-27B-Uncensored-Heretic-exl3-4.00bpw-h4-vb4 9
Total size
0 B
Files
2
Quantizations
1
Registered
2026-10-12 01:59
Last updated on HF
2026-10-12 01:35
Source huggingface.co/jirka642/Qwen3.8-27B-Uncensored-Heretic-Native-MTP-Preserved-exl3 ↗ Base model llmfan46/Qwen3.8-27B-Uncensored-Heretic-Native-MTP-Preserved →

Files by quantization

Auxiliary files 2 files 3.15 KB
.gitattributes 1.67 KB 0de2819e download
README.md 1.48 KB 7a6c0430 download

README current version from Hugging Face

base_model: llmfan46/Qwen3.8-27B-Uncensored-Heretic-Native-MTP-Preserved
base_model_relation: quantized
pipeline_tag: text-generation
library_name: exllamav3

EXL3 quants of llmfan46/Qwen3.8-27B-Uncensored-Heretic-Native-MTP-Preserved

  • Model Commit: 4b9e1517a43fdca25872318ed272a1c88e72c360
  • ExLlamaV3: 1.6.1
  • Command: python -m exllamav3.conversion.convert_model --in_dir "./llmfan46__Qwen3.8-27B-Uncensored-Heretic-Native-MTP-Preserved/" --out_dir "./4.0bpw_H6_V6" --work_dir "./workdir/" --bits 4 --mtp_bits 4 --head_bits 6 --vision_bits 6 --devices 0,1
  • llmfan46's repo lacks preprocessor_config.json, and the processor_config.json it has is not compatible, so I used preprocessor_config.json from Qwen/Qwen3.8-27B repo.

Weights:

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