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

SpaceTimee/Suri-Qwen-3.1-4B-Uncensored-Hard

SpaceTimee Qwen 4.0B
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/SpaceTimee%2FSuri-Qwen-3.1-4B-Uncensored-Hard"
Response includes
  • classification m-uncensored
  • files 13
  • benchmarks 11 entries
  • hub_downloads_all_time 173
  • providers 1
  • author_summary 31 models
  • readme_text full
Try next
GET/models?author=SpaceTimee
All 31 models by SpaceTimee
 GET/models?family=qwen&method=m-uncensored
Sibling qwen models · same method
 GET/methods/m-uncensored
What "m-uncensored" means · taxonomy
10 credits · hourly refresh · ~4 KB payload Get an API key →
Abliteration classifier · v1.0.0
M-U
Primary method

Uncensored (method unknown)

No other method signals detected in this model.
Confidence
LOW
Why this label 3 signals
Weak or ambiguous signals. Best guess based on catalog patterns; treat as tentative and check the evidence below.
  • 'uncensored' in name/tags but no 'abliterated' marker
  • method not identifiable from author declaration alone
  • may be DPO fine-tune, prompt engineering, or unknown technique
How we label → About M-U →
Refusal direction extraction

No specific extraction method could be identified for this model. The producer either did not document it or used a proprietary pipeline.

What is a refusal direction? →
Downloads · lifetime
173
31 last 30d - stable
Likes
1
Descendants
3
in 3 direct forks
Model age
10mo ago
created 2025-11-29
Available via
1 provider
featherless-ai

Training datasets

2 of 3 in /datasets

Corpora the author lists in the model card. Datasets tracked in our /datasets catalog carry a category badge linking to the workflow stage. Others open on Hugging Face.

NobodyExistsOnTheInternet/ToxicQAFinal Orion-zhen/dpo-toxic-zh heal unalignment/toxic-dpo-v0.2 heal
Downloads over time
Now183→from8↑2,188%
0671342018 on Dec 3, 2025183 on Oct 11Dec '25FebAprJunAugOct
Dec 3, 2025 → Oct 11 · 84 snapshots · spans 312 days

Benchmarks

Portrait before abliteration
Benchmarks of the base model as it stood before the refusal-removal operation. Compare with the numbers above to see what the operation cost.
Benchmark Score Source
Entertainment 0.4 UGI
Hazardous 1.2 UGI
Natural Intelligence 13.76 UGI
Political lean -12.4% UGI
Sensitive-Info 6.25 UGI
SocPol 0.5 UGI
UGI 15.83 UGI
Willingness (10) 3.5 UGI
W10-Adherence 1 UGI
W10-Direct 6 UGI
Writing 29.92 UGI

Genealogy 3 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.

Qwen/Qwen3-4B-Instruct-2507 this lineage
SpaceTimee/Suri-Qwen-3.1-4B-Uncensored-Hard↓ 31 3 forks
mradermacher/Suri-Qwen-3.1-4B-Uncensored-Hard-i1-GGUF↓ 1,326
mradermacher/Suri-Qwen-3.1-4B-Uncensored-Hard-GGUF↓ 323
SpaceTimee/Suri-Qwen-3.1-4B-Uncensored-Hard-GGUF↓ 245

Variants by this author 2 formats · 276 downloads combined

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

Metadata

Languages
zh en
Tags
transformers safetensors qwen3 text-generation suri qwen 4B uncensored unaligned text-generation-inference conversational zh

Related

Other by SpaceTimee
Suri-Qwen-3.8-27B-Uncensored-GGUF 13K Suri-Qwen-3.5-4B-Uncensored-GGUF 1K Suri-Qwen-3.5-9B-Uncensored-GGUF 1K Suri-Qwen-3.5-27B-Uncensored-GGUF 883 Suri-Qwen-3.8-27B-Uncensored-MLX 777
Same base model
HauhauCS/Qwen3-4B-2507-Instruct-Uncensored-HauhauCS-Aggressive 6K Unrestricted/Qwen3-4B-2507-Instruct-Uncensored-HauhauCS-Aggressive 3K huihui-ai/Huihui-Qwen3-4B-Instruct-2507-abliterated 2K IamLucif3r/Qwen3-4B-Instruct-2507-Abliterated 937 backpropSukuna/Qwen3-4B-Instruct-Uncensored 410
Total size
7.49 GB
Files
13
Quantizations
1
Registered
2026-08-22 13:56
Last updated on HF
2026-03-07 08:35
Source huggingface.co/SpaceTimee/Suri-Qwen-3.1-4B-Uncensored-Hard ↗ Base model Qwen/Qwen3-4B-Instruct-2507 ↗

Files by quantization

Auxiliary files 13 files 7.51 GB
model-00001-of-00002.safetensors 4.63 GB 05b003ca download
model-00002-of-00002.safetensors 2.87 GB 623cf2bd download
tokenizer.json 10.9 MB aeb13307 download
vocab.json 2.65 MB 4783fe10 download
merges.txt 1.59 MB 31349551 download
model.safetensors.index.json 32.1 KB b65d8063 download
tokenizer_config.json 9.43 KB 3ebf0b76 download
chat_template.jinja 3.95 KB a18870ad download
config.json 1.77 KB f2f703bc download
.gitattributes 1.53 KB 52373fe2 download
README.md 1.01 KB 19629080 download
added_tokens.json 707 B b54f9135 download
special_tokens_map.json 614 B 9b8043f1 download

README current version from Hugging Face

datasets:

  • NobodyExistsOnTheInternet/ToxicQAFinal
  • Orion-zhen/dpo-toxic-zh
  • unalignment/toxic-dpo-v0.2
    language:
  • zh
  • en
    base_model:
  • Qwen/Qwen3-4B-Instruct-2507
  • unsloth/Qwen3-4B-Instruct-2507-unsloth-bnb-4bit
    tags:
  • suri
  • qwen
  • 4B
  • uncensored
  • unaligned
  • text-generation-inference
    pipeline_tag: text-generation
    library_name: transformers

Suri ⬡ Qwen

Uncensor Is All You Need


自我介绍

Suri Qwen 3.1 4B Uncensored Hard: 一只基于 Qwen3 4B Instruct 2507 的无审查 (去对齐) 模型

与主模型 Suri Qwen 3.1 4B Uncensored 相比，该模型使用更大学习率训练，去对齐效果更强，语言灵活性更差，请仅在主模型拒绝回答时才使用该模型

推荐参数

  • Temperature: 0.5 ~ 0.7
  • Top_P: 0.8 ~ 0.95
  • Presence_Penalty: 0 ~ 2

开发者

Space Time

联系方式

  1. QQ 群: 902575634
  2. 邮箱: [email protected]

•ᴗ•

README history 3 versions

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

  1. 2026-03-07Update README.md0890fda1 KB
    Loading...
  2. 2025-11-29Update README.mdebb3940914 B
    Loading...
  3. 2025-11-29Unsloth Model Card159d72c621 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