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

Naphula/Darkest-muse-v1-abliterated

Naphula Gemma 10B
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/Naphula%2FDarkest-muse-v1-abliterated"
Response includes
  • classification m1
  • files 13
  • benchmarks 11 entries
  • hub_downloads_all_time 157
  • author_summary 22 models
  • readme_text full
Try next
GET/models?author=Naphula
All 22 models by Naphula
 GET/models?family=gemma&method=m1
Sibling gemma models · same method
 GET/methods/m1
What "m1" means · taxonomy
10 credits · hourly refresh · ~4 KB payload Get an API key →
Abliteration classifier · v1.0.0
M1
Primary method

Direct removal

No other method signals detected in this model.
Confidence
MEDIUM
Why this label 3 signals
Method inferred from partial signals - repository name, related files, or tag patterns. Producer identity not confirmed; label may sharpen or shift as we gather more evidence.
  • 'abliterated' in name/tags
  • is_gguf=0 (base model)
  • no specific method indicators - defaulting to M1 (most common)
How we label → About M1 →
Refusal direction extracted via
Extraction technique

Difference-of-means

FromArditi et al. (2024) ↗
Confidence
MEDIUM
Why we say so
primary_method=M1; difference-of-means is the reference extraction for M1/M3 (Arditi 2024)
What is a refusal direction? → Read the paper ↗
Downloads · lifetime
157
38 last 30d - stable
Likes
2
Descendants
2
in 2 direct forks
Model age
10mo ago
created 2025-12-08

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.

sam-paech/gutenberg3-generalfiction-scifi-fantasy-romance-adventure-dpo
Downloads over time
Now171→from42↑307%
368513418442 on Dec 10, 2025171 on Oct 11171 on Oct 8Dec '25FebAprJunAugOct
Dec 10, 2025 → Oct 11 · 83 snapshots · spans 305 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 0.8 UGI
Hazardous 1.8 UGI
Natural Intelligence 13.17 UGI
Political lean -23.5% UGI
Sensitive-Info 9.64 UGI
SocPol 0.6 UGI
UGI 11.42 UGI
Willingness (10) 1.5 UGI
W10-Adherence 1 UGI
W10-Direct 2 UGI
Writing 20.03 UGI

Genealogy 2 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.

sam-paech/Darkest-muse-v1 this lineage
Naphula/Darkest-muse-v1-abliterated↓ 38 2 forks
mradermacher/Darkest-muse-v1-abliterated-i1-GGUF↓ 1,200
mradermacher/Darkest-muse-v1-abliterated-GGUF↓ 500

Variants by this author 2 formats · 77 downloads combined

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

Metadata

License
gemma
Tags
transformers safetensors gemma2 text-generation creative creative-writing conversational dataset:sam-paech/gutenberg3-generalfiction-scifi-fantasy-romance-adventure-dpo base_model:sam-paech/Darkest-muse-v1 base_model:finetune:sam-paech/Darkest-muse-v1 license:gemma text-generation-inference

Related

Other by Naphula
GLM-4.5-Air-Abliterated-FQ3_K_XL-GGUF 430 magnum-v4-9b-abliterated-GGUF 233 MN-12B-Mag-Mell-R1-Uncensored-GGUF 108 MN-12B-Mag-Mell-R1-Uncensored 106 gemma-3-27b-it-abliterated-Q8_K_XL-GGUF 95
Same base model
Naphula/Darkest-muse-v1-abliterated-GGUF 39
Total size
18.9 GB
Files
13
Quantizations
1
Registered
2026-08-22 13:56
Last updated on HF
2025-12-08 16:25
Source huggingface.co/Naphula/Darkest-muse-v1-abliterated ↗ Base model sam-paech/Darkest-muse-v1 ↗

Files by quantization

Auxiliary files 13 files 18.9 GB
model-00002-of-00005.safetensors 4.64 GB a2f7e754 download
model-00004-of-00005.safetensors 4.64 GB 5494464e download
model-00001-of-00005.safetensors 4.62 GB fc25a1bf download
model-00003-of-00005.safetensors 4.59 GB 7401072f download
model-00005-of-00005.safetensors 448 MB 0c57712c download
tokenizer.json 16.7 MB 3f289bc0 download
tokenizer.model 4.04 MB 61a7b147 download
tokenizer_config.json 45.9 KB 9b18de69 download
model.safetensors.index.json 36.4 KB 5dbdc05f download
.gitattributes 1.53 KB 52373fe2 download
README.md 1.09 KB 4d2f505a download
config.json 906 B d94f80a9 download
special_tokens_map.json 636 B 8d6368f7 download

README current version from Hugging Face

library_name: transformers
license: gemma
base_model: sam-paech/Darkest-muse-v1
tags:

uploading these safetensors before they get deleted.

they have not been ablated with MPOA, so may suffer some cognitive decline.

made these mainly for testing purpose, to see what it might look like in the psycho merge.

i did not measure the compliance score of this (untested to see if it produce bugged output)

i recommend not using these in a merge. better to merge the models and then ablate afterward with https://huggingface.co/blog/grimjim/norm-preserving-biprojected-abliteration

README history 3 versions

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

  1. 2025-12-08Update README.md99c90f51.1 KB
    Loading...
  2. 2025-12-08Update README.md187063a491 B
    Loading...
  3. 2025-12-08Create README.mdeb8178f397 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