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

nuofang/Qwen3.5-9B-ultra-uncensored-heretic-v2-GGUF

nuofang Qwen 9B GGUF second-order 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/nuofang%2FQwen3.5-9B-ultra-uncensored-heretic-v2-GGUF"
Response includes
  • classification m3
  • files 8
  • hub_downloads_all_time 4,046
  • author_summary 22 models
  • readme_text full
Try next
GET/models?author=nuofang
All 22 models by nuofang
 GET/models?family=qwen&method=m3
Sibling qwen 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 · lifetime
4K
277 last 30d - cooling
Likes
1
Model age
3mo ago
created 2026-06-27
Downloads over time
Now4.2K→from0↑0%
01.5K3.1K4.6K0 on Jun 244.2K on Oct 12JunJulAugSepOct
Jun 24 → Oct 12 · 56 snapshots · spans 110 days

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/Qwen3.5-9B-ultra-uncensored-heretic-v2 this lineage
nuofang/Qwen3.5-9B-ultra-uncensored-heretic-v2-GGUF↓ 277

Metadata

Quantizations
IQ4 Q4_K
Tags
gguf llama.cpp quantized imatrix base_model:llmfan46/Qwen3.5-9B-ultra-uncensored-heretic base_model:quantized:llmfan46/Qwen3.5-9B-ultra-uncensored-heretic endpoints_compatible region:us conversational

Related

Other by nuofang
Qwen3.8-9B-Distill-SLERP-F451-Pro-Writer-Uncensored-GGUF 15K Qwen3.8-9B-Distill-uncensored-heretic-GGUF 4K Huihui-Qwen3.5-9B-abliterated-GGUF 3K Qwen3.8-9B-abliterated-25-GGUF 2K Qwen3.5-9B-The-Bradbury-F451-Pro-Writer-Uncensored-Heretic-GGUF 993
Same base model
marafx2025/Qwen3.5-9B-ultra-uncensored-heretic-v2-i1-GGUF 672 DBMe/Qwen3.5-9B-ultra-uncensored-heretic-v2-exl3 33
Total size
20.1 GB
Files
8
Quantizations
4
Registered
2026-08-22 13:56
Last updated on HF
2026-07-05 22:07
Source huggingface.co/nuofang/Qwen3.5-9B-ultra-uncensored-heretic-v2-GGUF ↗ Base model llmfan46/Qwen3.5-9B-ultra-uncensored-heretic-v2 ↗

Files by quantization

Q4_K 2 files 10.2 GB
Qwen3.5-9B-ultra-uncensored-heretic-v2-Q4_K_M.gguf 5.24 GB 5548413c download
Qwen3.5-9B-ultra-uncensored-heretic-v2-Q4_K_S.gguf 4.98 GB dc00a112 download
IQ4 2 files 9.89 GB
Qwen3.5-9B-ultra-uncensored-heretic-v2-IQ4_NL.gguf 5.05 GB 51b97d13 download
Qwen3.5-9B-ultra-uncensored-heretic-v2-IQ4_XS.gguf 4.84 GB 694fbe7b download
F16 1 file 960 B
mmproj-Qwen3.5-9B-ultra-uncensored-heretic-v2-f16.gguf 960 B 2adb1c29 download
Auxiliary files 3 files 4.91 MB
imatrix.gguf 4.91 MB 3d29447b download
.gitattributes 1.87 KB 994773b3 download
README.md 813 B 874e0f60 download

README current version from Hugging Face

base_model: llmfan46/Qwen3.5-9B-ultra-uncensored-heretic-v2
tags:

  • llama.cpp
  • quantized
  • imatrix

Auto-Quantized GGUF Model

This repository contains automated GGUF quantization files for llmfan46/Qwen3.5-9B-ultra-uncensored-heretic-v2.

The calibration data for the imatrix is targeted at Chinese novels and role-playing (RP), while preserving logic and common sense.
imatrix 的校准数据以中文的小说、角色扮演为目标，同时保留逻辑和常识。

📊 Perplexity Evaluation

(Tested against the provided calibration dataset)

  • Base (F16/BF16): PPL = 14.2066 +/- 0.11553
  • IQ4_XS: PPL = 12.0006 +/- 0.09439
  • IQ4_NL: PPL = 12.0064 +/- 0.09447
  • Q4_K_M: PPL = 11.9559 +/- 0.09373

README history 10 versions

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

  1. 2026-07-05Upload README.md with huggingface_hub33ab56d813 B
    Loading...
  2. 2026-07-05Upload README.md with huggingface_hubeb8fb88773 B
    Loading...
  3. 2026-07-05Upload README.md with huggingface_hubd8c1096733 B
    Loading...
  4. 2026-07-05Upload README.md with huggingface_hub9d4d66b693 B
    Loading...
  5. 2026-07-05Upload README.md with huggingface_hubf094495559 B
    Loading...
  6. 2026-06-28Upload README.md with huggingface_hubedc6249684 B
    Loading...
  7. 2026-06-28Upload README.md with huggingface_hub9bc3b95559 B
    Loading...
  8. 2026-06-27Upload README.md with huggingface_hub22d5496733 B
    Loading...
  9. 2026-06-27Upload README.md with huggingface_hub7e96f5f693 B
    Loading...
  10. 2026-06-27Upload README.md with huggingface_hub7aa2d9d559 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