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

armand0e/Qwen3.5-9B-ultra-uncensored-heretic

armand0e Qwen 9.4B multimodal 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/armand0e%2FQwen3.5-9B-ultra-uncensored-heretic"
Response includes
  • classification m3
  • files 10
  • hub_downloads_all_time 1,372
  • author_summary 1 models
  • readme_text full
Try next
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
1K
965 last 30d - active
Likes
2
Model age
2mo ago
created 2026-07-28
Downloads over time
Now1.8K→from31↑5,665%
06541.3K2K31 on Jul 291.8K on Oct 11JulAugSepOct
Jul 29 → Oct 11 · 51 snapshots · spans 74 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↓ 512 this lineage
armand0e/Qwen3.5-9B-ultra-uncensored-heretic↓ 965

Metadata

License
apache-2.0
Languages
en
Tags
transformers safetensors qwen3_5 image-text-to-text repaired-checkpoint conversational en base_model:llmfan46/Qwen3.5-9B-ultra-uncensored-heretic base_model:finetune:llmfan46/Qwen3.5-9B-ultra-uncensored-heretic license:apache-2.0 endpoints_compatible region:us

Related

Same base model
llmfan46/Qwen3.5-9B-ultra-uncensored-heretic-GGUF 8K mradermacher/Qwen3.5-9B-ultra-uncensored-heretic-v2-i1-GGUF 3K mradermacher/Qwen3.5-9B-ultra-uncensored-heretic-v2-GGUF 2K colddRazer/Qwen3.5-9B-ultra-uncensored-heretic-EXL3-4bpw 34
Total size
17.5 GB
Files
10
Quantizations
1
Registered
2026-08-22 13:56
Last updated on HF
2026-07-28 15:28
Source huggingface.co/armand0e/Qwen3.5-9B-ultra-uncensored-heretic ↗ Base model llmfan46/Qwen3.5-9B-ultra-uncensored-heretic →

Files by quantization

Auxiliary files 10 files 17.5 GB
model.safetensors 17.5 GB 0d787f5b download
tokenizer.json 19.1 MB 639e352c download
chat_template.jinja 7.83 KB 1e6f3b77 download
config.json 2.86 KB 1853edd0 download
.gitattributes 1.53 KB 52373fe2 download
tokenizer_config.json 1.10 KB b1f21853 download
README.md 618 B ededcbf2 download
preprocessor_config.json 390 B 2ea84a43 download
video_preprocessor_config.json 385 B 3ba673a5 download
generation_config.json 121 B 4658fb98 download

README current version from Hugging Face

language:

  • en
    license: apache-2.0
    library_name: transformers
    pipeline_tag: image-text-to-text
    tags:
  • qwen3_5
  • repaired-checkpoint
    base_model: llmfan46/Qwen3.5-9B-ultra-uncensored-heretic

Qwen3.5-9B Ultra Uncensored Heretic (repaired)

This is a Transformers-compatible repair of the upstream Qwen3.5 checkpoint.

The original safetensors stored 759 tensors under invalid prefixes: 426 language-model tensors under a triple language_model nesting and 333 vision tensors under model.language_model.visual. This release changes safetensors header names only; raw tensor data is preserved byte-for-byte.

README history 2 versions

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

  1. 2026-07-28Update README.mdf56bbf5618 B
    Loading...
  2. 2026-07-28Publish repaired Qwen3.5 base checkpoint01d76c8700 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