Get the app
← back to catalog · registered 2026-11-06 23:02

tstepspam/Qwen3.8-27B-Ultimate-Heretic-Uncensored-Q4-MLX

tstepspam 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/tstepspam%2FQwen3.8-27B-Ultimate-Heretic-Uncensored-Q4-MLX"
Response includes
  • classification m3
  • files 11
  • author_summary 8 models
  • readme_text full
Try next
GET/models?author=tstepspam
All 8 models by tstepspam
 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 2026-11-06

Training 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.

DavidAU/Polar-STRICT-Datasets DavidAU/F451-STRICT-Datasets
Downloads over time
Now0→from0↑0%
00110 on Nov 60 on Nov 7Nov
Nov 6 → Nov 7 · 2 snapshots · spans 1 day

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.

DavidAU/Qwen3.8-27B-Ultimate-Heretic-Uncensored↓ 803 this lineage
tstepspam/Qwen3.8-27B-Ultimate-Heretic-Uncensored-Q4-MLX

Variants by this author 2 formats · 0 downloads combined

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

Metadata

License
apache-2.0
Languages
en
Tags
mlx safetensors qwen3_5 heretic qwen3_8 qwen3_6 uncensored finetune Cold Fusion GAIN Training Multi-stage tuning all use cases

Related

Other by tstepspam
Huihui-Qwen3.5-35B-A3B-Claude-4.6-Opus-abliterated-Q8-MLX 442 Huihui-Qwen3.8-27B-abliterated-Q4-MLX 367 Huihui-Qwen3.5-35B-A3B-Claude-4.6-Opus-abliterated-Q4-MLX 312 Huihui-Qwen3.8-27B-abliterated-Q8-MLX 255 Huihui-Qwen3.5-27B-abliterated-Q8-MLX 210
Same base model
DavidAU/Qwen3.8-27B-Ultimate-Heretic-Uncensored-Turbo-Brilliance-50x-AGI-NEO-GGUF 10K mradermacher/Qwen3.8-27B-Ultimate-Heretic-Uncensored-GGUF 859 TheUnderscore/Qwen3.8-27B-Ultimate-Heretic-Uncensored-NVFP4-W4ATTN-AWQ 50 TheUnderscore/Qwen3.8-27B-Ultimate-Heretic-Uncensored-W4A16-AWQ 0 tstepspam/Qwen3.8-27B-Ultimate-Heretic-Uncensored-Q8-MLX 0
Total size
14.1 GB
Files
11
Quantizations
1
Registered
2026-11-06 23:02
Last updated on HF
2026-11-06 22:08
Source huggingface.co/tstepspam/Qwen3.8-27B-Ultimate-Heretic-Uncensored-Q4-MLX ↗ Base model DavidAU/Qwen3.8-27B-Ultimate-Heretic-Uncensored →

Files by quantization

Auxiliary files 11 files 14.1 GB
model-00002-of-00003.safetensors 4.99 GB 1d43b2eb download
model-00001-of-00003.safetensors 4.96 GB 8e785524 download
model-00003-of-00003.safetensors 4.14 GB 5f8e0a48 download
tokenizer.json 19.1 MB 06b95093 download
chat_template.jinja 242 KB 9f11ce44 download
model.safetensors.index.json 185 KB 8cabf784 download
config.json 4.09 KB c71f9713 download
.gitattributes 1.53 KB 52373fe2 download
tokenizer_config.json 1.24 KB 0b282af8 download
README.md 811 B ed5dbcf5 download
generation_config.json 214 B 0bc3addd download

README current version from Hugging Face

license: apache-2.0
language:

  • en
    pipeline_tag: text-generation
    library_name: mlx
    tags:
  • heretic
  • qwen3_8
  • qwen3_6
  • uncensored
  • finetune
  • Cold Fusion
  • GAIN Training
  • Multi-stage tuning
  • all use cases
  • unsloth
  • 27 reasoning modes
  • 27 instruct modes
  • 15 AI Constructs
  • 5 Super AI Constructs
  • 21 AGI Constructs
  • Brainstorm AI Construct
  • Brainstorm Business AI Construct
  • Universal Teacher AI Construct
  • Agent AI Construct
  • Coder AI Construct
  • Turbo-Brilliance
  • 9 Xhigh reasoning modes
  • fine tune
  • multi-stage tuned
  • thinking
  • reasoning
  • coder
  • creative
  • creative writing
  • all genres
  • story
  • writing
  • fiction
  • roleplaying
  • bfloat16
  • mlx
    datasets:
  • DavidAU/Polar-STRICT-Datasets
  • DavidAU/F451-STRICT-Datasets
    base_model: DavidAU/Qwen3.8-27B-Ultimate-Heretic-Uncensored
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