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

kabachuha/gemma3-4b-it-abliterated

kabachuha Gemma 3.3B multimodal
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/kabachuha%2Fgemma3-4b-it-abliterated"
Response includes
  • classification m1
  • files 15
  • benchmarks 16 entries
  • hub_downloads_all_time 463
  • author_summary 2 models
  • readme_text full
Try next
GET/models?author=kabachuha
All 2 models by kabachuha
 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
463
16 last 30d - cooling
Likes
1
Descendants
2
in 2 direct forks
Model age
19mo ago
created 2025-03-16
Downloads over time
Now468→from89↑426%
7021536150689 on Mar 12, 2025468 on Oct 11Mar '25Jun '25Sep '25Dec '25MarJunSep
Mar 12, 2025 → Oct 11 · 122 snapshots · spans 578 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
Arena-Battles 4321 LM-Arena
LM Arena Elo 1293.3101140666447 LM-Arena
Arena-Elo-Lower 1284.2314692781533 LM-Arena
Arena-Elo-Upper 1302.388758855136 LM-Arena
Arena-Rank 77 LM-Arena
Entertainment 1.3 UGI
Hazardous 1.2 UGI
Natural Intelligence 11.19 UGI
Political lean -19.3% UGI
Sensitive-Info 10.35 UGI
SocPol 0.6 UGI
UGI 16.9 UGI
Willingness (10) 3 UGI
W10-Adherence 1 UGI
W10-Direct 5 UGI
Writing 24.01 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.

google/gemma-3-4b-it this lineage
kabachuha/gemma3-4b-it-abliterated↓ 16 2 forks
mradermacher/gemma3-4b-it-abliterated-i1-GGUF↓ 1,070
mradermacher/gemma3-4b-it-abliterated-GGUF↓ 695

Metadata

License
gemma
Tags
transformers safetensors gemma3 image-text-to-text conversational base_model:google/gemma-3-4b-it base_model:finetune:google/gemma-3-4b-it license:gemma text-generation-inference endpoints_compatible region:us

Related

Other by kabachuha
Marco-o1-abliterated-Q8_0-GGUF 22
Same base model
mlabonne/gemma-3-4b-it-abliterated-GGUF 2K matrixportalx/gemma-3-4b-it-abliterated-GGUF 1K arzaan789/gemma-3-4b-uncensored 435 arzaan789/gemma-3-4b-uncensored_ 339 jwest33/gemma-3-4b-it-null-space-abliterated-GGUF 324
Total size
8.01 GB
Files
15
Quantizations
1
Registered
2026-08-22 13:56
Last updated on HF
2025-03-16 11:01
Source huggingface.co/kabachuha/gemma3-4b-it-abliterated ↗ Base model google/gemma-3-4b-it ↗

Files by quantization

Auxiliary files 15 files 8.05 GB
model-00001-of-00002.safetensors 4.62 GB d853041f download
model-00002-of-00002.safetensors 3.39 GB f024a108 download
tokenizer.json 31.8 MB 4667f208 download
tokenizer.model 4.47 MB 1299c11d download
tokenizer_config.json 1.10 MB 7bdd14f0 download
model.safetensors.index.json 88.4 KB 4b95241f download
.gitattributes 1.72 KB e6d43da7 download
chat_template.json 1.58 KB 719b0cd0 download
config.json 1.58 KB 9fdf5e85 download
special_tokens_map.json 662 B 1a619324 download
preprocessor_config.json 570 B b1e00fc1 download
README.md 550 B 256a59d4 download
generation_config.json 192 B f60a6730 download
processor_config.json 70.0 B 453c7966 download
added_tokens.json 35.0 B e17bde03 download

README current version from Hugging Face

license: gemma
library_name: transformers
pipeline_tag: image-text-to-text
base_model: google/gemma-3-4b-it

Gemma 3 abliterated

This is Gemma 3 4b with [abliteration technique] (https://www.lesswrong.com/posts/jGuXSZgv6qfdhMCuJ/refusal-in-llms-is-mediated-by-a-single-direction) applied to reduce refusals. (with high scale_factor=2.0 as it was very stubborn originally)

Tested its work in VLLM. It does give the disclaimers sometimes, but more or less often executes the instructions as needed, and the disclaimers are less aggressive.

README history 1 version

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

  1. 2025-03-16Upload folder using huggingface_hub17f70c0550 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