← back to catalog · registered 2026-08-31 12:03

mradermacher/Tencent-Hy-30B-A3B-uncensored-heretic-GGUF

mradermacher 30B MoE 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/mradermacher%2FTencent-Hy-30B-A3B-uncensored-heretic-GGUF"
Response includes
  • classification m3
  • files 3
  • hub_downloads_all_time 988
  • author_summary 3364 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 · lifetime
988
331 last 30d - stable
Likes
0
Model age
7w ago
created 2026-08-31
Downloads over time
Now1.2K→from0↑0%
04358691.3K0 on Aug 311.2K on Oct 201.2K on Oct 19AugSepOct
Aug 31 → Oct 20 · 51 snapshots · spans 50 days

Variants by this author 2 formats · 1K downloads combined

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

Metadata

Quantizations
Q4_K
Tags
transformers gguf translation heretic uncensored decensored abliterated ara-lora zh en fr pt

Related

Total size
16.0 GB
Files
3
Quantizations
2
Registered
2026-08-31 12:03
Last updated on HF
2026-08-31 17:35

Files by quantization

Q4_K 1 file 16.0 GB
Tencent-Hy-30B-A3B-uncensored-heretic.Q4_K_S.gguf 16.0 GB 9c84d271 download
Auxiliary files 2 files 1.95 KB
.gitattributes 1.57 KB 3e26ae88 download
README.md 391 B 558b2fd2 download

README current version from Hugging Face

README history 7 versions

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

  1. 2026-08-31auto-patch README.mdcd5b6d64.1 KB
    Loading...
  2. 2026-08-31auto-patch README.md81406953.9 KB
    Loading...
  3. 2026-08-31auto-patch README.md899d43a3.9 KB
    Loading...
  4. 2026-08-31auto-patch README.md75408e53.5 KB
    Loading...
  5. 2026-08-31auto-patch README.md148d2553.2 KB
    Loading...
  6. 2026-08-31auto-patch README.md7471c042.6 KB
    Loading...
  7. 2026-08-31uploaded from rich1caf5ebc391 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