Get the app
← back to catalog · registered 2026-10-17 04:59

mradermacher/LFM2.5-24B-A2B-Uncensored-Heretic-i1-GGUF

mradermacher 24B GGUF
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%2FLFM2.5-24B-A2B-Uncensored-Heretic-i1-GGUF"
Response includes
  • classification m3
  • files 7
  • author_summary 3353 models
  • readme_text full
Try next
GET/models?author=mradermacher
All 3353 models by mradermacher
 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 2026-10-17

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

Quantizations
IQ3 Q2_K Q4_K
Tags
transformers gguf liquid lfm2.5 moe heretic uncensored decensored abliterated en ar zh

Related

Other by mradermacher
Ornith-1.5-9B-uncensored-GGUF 1.7M Qwen3-VL-8B-Instruct-abliterated-GGUF 664K Qwen3.8-Flash-Next-Uncensored-i1-GGUF 157K gemma-4-26B-A4B-it-ultra-uncensored-heretic-i1-GGUF 119K Qwen3.8-9B-heretic-uncensored-GGUF 113K
Total size
39.5 GB
Files
7
Quantizations
4
Registered
2026-10-17 04:59
Last updated on HF
2026-10-17 05:51
Source huggingface.co/mradermacher/LFM2.5-24B-A2B-Uncensored-Heretic-i1-GGUF ↗

Files by quantization

Q4_K 1 file 12.7 GB
LFM2.5-24B-A2B-Uncensored-Heretic.i1-Q4_K_S.gguf 12.7 GB e3b855c1 download
IQ3 2 files 18.4 GB
LFM2.5-24B-A2B-Uncensored-Heretic.i1-IQ3_M.gguf 9.80 GB 719d84dd download
LFM2.5-24B-A2B-Uncensored-Heretic.i1-IQ3_XXS.gguf 8.65 GB 9e5dd73e download
Q2_K 1 file 8.20 GB
LFM2.5-24B-A2B-Uncensored-Heretic.i1-Q2_K.gguf 8.20 GB ba199483 download
Auxiliary files 3 files 107 MB
LFM2.5-24B-A2B-Uncensored-Heretic.imatrix.gguf 107 MB cd9d7811 download
.gitattributes 1.89 KB 421e357d download
README.md 491 B acfd6ef3 download

README current version from Hugging Face

README history 2 versions

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

  1. 2026-10-17auto-patch README.md0ab99223.9 KB
    Loading...
  2. 2026-10-17uploaded from marcob3c6d62491 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