← back to catalog · registered 2026-11-20 05:04

LTM-Lab/Qwen3.8-40B-Ultimate-Heretic-Uncensored-Build2-v16

LTM-Lab 40B 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/LTM-Lab%2FQwen3.8-40B-Ultimate-Heretic-Uncensored-Build2-v16"
Response includes
  • classification m3
  • files 31
  • author_summary 17 models
  • readme_text full
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
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)
Refusal direction extracted via
Extraction technique

Difference-of-means

Confidence
HIGH
Why we say so
name contains 'heretic'; Heretic default extraction is difference-of-means (Arditi 2024)
Downloads · 30-day
38
Likes
0
Descendants
1
in 1 direct fork
Model age
2d ago
created 2026-11-20
Downloads over time
Now 38 → from 0 ↑ 0%
0142842 0 on Nov 20 38 on Nov 22 Nov
Nov 20 → Nov 22 · 3 snapshots · spans 2 days

Genealogy 1 direct fork

Full fork graph →

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

Metadata

License
apache-2.0
Tags
transformers safetensors qwen3_5 image-text-to-text v16-episodic-memory sglang heretic qwen3_8 qwen3_6 uncensored abliterated finetune

Related

Total size
74.4 GB
Files
31
Quantizations
1
Registered
2026-11-20 05:04
Last updated on HF
2026-11-20 11:37

Files by quantization

Auxiliary files 31 files 74.5 GB
model-00014-of-00017.safetensors 4.65 GB 63bf4e21 download
model-00011-of-00017.safetensors 4.62 GB ad5f4dba download
model-00003-of-00017.safetensors 4.62 GB 095eb614 download
model-00007-of-00017.safetensors 4.61 GB 2b1583cd download
model-00008-of-00017.safetensors 4.60 GB 2eb0f06e download
model-00015-of-00017.safetensors 4.59 GB a3167b23 download
model-00009-of-00017.safetensors 4.59 GB a5f1a3f5 download
model-00004-of-00017.safetensors 4.58 GB 85699dc8 download
model-00012-of-00017.safetensors 4.57 GB 484a37c8 download
model-00006-of-00017.safetensors 4.57 GB e11f3912 download
model-00016-of-00017.safetensors 4.57 GB f3e4765b download
model-00013-of-00017.safetensors 4.52 GB 32d600d9 download
model-00002-of-00017.safetensors 4.51 GB e0db98c8 download
model-00010-of-00017.safetensors 4.51 GB 7c21ca00 download
model-00005-of-00017.safetensors 4.50 GB 6089fa37 download
model-00017-of-00017.safetensors 3.44 GB 63c08eec download
model-00001-of-00017.safetensors 2.37 GB 54d83c1d download
tokenizer.json 19.1 MB 6f32ce20 download
vocab.json 6.41 MB 0aa0ce06 download
modeling_qwen_dense_v16.py 340 KB 76565f31 download
chat_template.jinja 245 KB 074c8ad4 download
model.safetensors.index.json 152 KB 019d20aa download
configuration_qwen_dense_v16.py 10.3 KB adecf415 download
tokenizer_config.json 8.48 KB 1a7e62d0 download
config.json 5.36 KB c53a28f6 download
.gitattributes 1.53 KB 52373fe2 download
processor_config.json 1.16 KB 33818c7f download
README.md 570 B 9ceb3ca4 download
preprocessor_config.json 390 B 2ea84a43 download
video_preprocessor_config.json 385 B 3ba673a5 download
generation_config.json 213 B d04042de download

README current version from Hugging Face


license: apache-2.0
base_model: LTM-Lab/Qwen3.8-40B-Ultimate-Heretic-Uncensored-Build2-v15
tags:

  • v16-episodic-memory
  • transformers
    pipeline_tag: image-text-to-text

Qwen3.8-40B-Ultimate-Heretic-Uncensored-Build2-v16

This release retains the pretrained weights and tokenizer of the source checkpoint. The included Hugging Face architecture adds persistent V16 episodic memory at inference. Load with trust_remote_code=True.

README history 2 versions

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

  1. 2026-11-20 Expand model card with version-specific memory, serving and measurements 7f4306a 32.1 KB
    Loading...
  2. 2026-11-20 Add V16 model files ee2544a 570 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.

Open in Abliteration