← back to catalog · registered 2026-11-12 04:03

mradermacher/gemma-4-26B-A4B-it-abliterated-v3-i1-GGUF

mradermacher Gemma 26B GGUF MoE 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/mradermacher%2Fgemma-4-26B-A4B-it-abliterated-v3-i1-GGUF"
Response includes
  • classification m8
  • files 16
  • author_summary 3602 models
  • readme_text full
10 credits · hourly refresh · ~4 KB payload Get an API key →
Abliteration classifier · v1.0.0
M8
Primary method

Repackaging (quantization)

Applied on top of direct removal inherited from the base model.
Confidence
HIGH
Inherited from base model
Why this label 3 signals
Producer identity confirmed by naming conventions, tags or the model card. This label is very unlikely to change.
  • author=mradermacher (M8 quantization producer, never originator)
  • is_gguf=1
  • base_model='jsantillana/gemma-4-26B-A4B-it-abliterated-v3' (base has 'abliterated' marker, assume M1 default)
Refusal direction extraction

No specific extraction method could be identified for this model. The producer either did not document it or used a proprietary pipeline.

What is a refusal direction? →
Downloads · 30-day
7K
Likes
1
Model age
1w ago
created 2026-11-12
Downloads over time
Now 7.1K → from 0 ↑ 0%
02.6K5.2K7.8K 0 on Nov 12 7.1K on Nov 22 Nov
Nov 12 → Nov 22 · 11 snapshots · spans 10 days

Genealogy 0 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.

Variants by this author 2 formats · 9K downloads combined

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

Metadata

License
other
Languages
en es
Quantizations
IQ1 IQ2 IQ3 IQ4 Q2_K Q3_K Q4_K Q6_K
Tags
transformers gguf abliteration refusal-direction cybersecurity interpretability research gemma en es base_model:jsantillana/gemma-4-26B-A4B-it-abliterated-v3 base_model:quantized:jsantillana/gemma-4-26B-A4B-it-abliterated-v3

Related

Total size
156 GB
Files
16
Quantizations
10
Registered
2026-11-12 04:03
Last updated on HF
2026-11-12 05:09

Files by quantization

Q6_K 1 file 21.1 GB
gemma-4-26B-A4B-it-abliterated-v3.i1-Q6_K.gguf 21.1 GB 0692eca9 download
Q4_K 2 files 30.0 GB
gemma-4-26B-A4B-it-abliterated-v3.i1-Q4_K_M.gguf 15.6 GB 56951fd1 download
gemma-4-26B-A4B-it-abliterated-v3.i1-Q4_K_S.gguf 14.4 GB 33746691 download
IQ4 1 file 13.0 GB
gemma-4-26B-A4B-it-abliterated-v3.i1-IQ4_XS.gguf 13.0 GB 3b155dd5 download
Q3_K 2 files 23.8 GB
gemma-4-26B-A4B-it-abliterated-v3.i1-Q3_K_M.gguf 12.4 GB 0c3ec034 download
gemma-4-26B-A4B-it-abliterated-v3.i1-Q3_K_S.gguf 11.4 GB 3ba253da download
IQ3 2 files 22.1 GB
gemma-4-26B-A4B-it-abliterated-v3.i1-IQ3_M.gguf 11.5 GB ff422fb4 download
gemma-4-26B-A4B-it-abliterated-v3.i1-IQ3_XXS.gguf 10.5 GB 63b9fdd7 download
Q2_K 2 files 19.8 GB
gemma-4-26B-A4B-it-abliterated-v3.i1-Q2_K_S.gguf 9.89 GB d2b1351a download
gemma-4-26B-A4B-it-abliterated-v3.i1-Q2_K.gguf 9.86 GB 5f867327 download
IQ2 2 files 18.3 GB
gemma-4-26B-A4B-it-abliterated-v3.i1-IQ2_M.gguf 9.67 GB 666c9ab3 download
gemma-4-26B-A4B-it-abliterated-v3.i1-IQ2_XXS.gguf 8.66 GB 820a5c4f download
IQ1 1 file 8.07 GB
gemma-4-26B-A4B-it-abliterated-v3.i1-IQ1_M.gguf 8.07 GB 58cd96ea download
Drafts and companions 1 file 109 MB
gemma-4-26B-A4B-it-abliterated-v3.imatrix.gguf 109 MB bf809c4c download
Auxiliary files 2 files 6.83 KB
README.md 4.20 KB 2dc98a17 download
.gitattributes 2.64 KB b7cfe4fd download

README current version from Hugging Face


base_model: jsantillana/gemma-4-26B-A4B-it-abliterated-v3
language:

  • en
  • es
    library_name: transformers
    license: other
    license_name: apache-2.0-restricted-research
    mradermacher:
    readme_rev: 1
    quantized_by: mradermacher
    tags:
  • abliteration
  • refusal-direction
  • cybersecurity
  • interpretability
  • research
  • gemma

About

weighted/imatrix quants of https://huggingface.co/jsantillana/gemma-4-26B-A4B-it-abliterated-v3

For a convenient overview and download list, visit our model page for this model.

static quants are available at https://huggingface.co/mradermacher/gemma-4-26B-A4B-it-abliterated-v3-GGUF

This is a vision model - mmproj files (if any) will be in the static repository.

Usage

If you are unsure how to use GGUF files, refer to one of TheBloke's
READMEs
for
more details, including on how to concatenate multi-part files.

Provided Quants

(sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)

Link Type Size/GB Notes
GGUF imatrix 0.2 imatrix file (for creating your own quants)
GGUF i1-IQ2_M 10.5
GGUF i1-Q2_K 10.7 IQ3_XXS probably better
GGUF i1-IQ3_XXS 11.4 lower quality
GGUF i1-IQ3_M 12.5
GGUF i1-Q3_K_M 13.4 IQ3_S probably better
GGUF i1-Q4_K_S 15.6 optimal size/speed/quality
GGUF i1-Q4_K_M 16.9 fast, recommended
GGUF i1-Q6_K 22.7 practically like static Q6_K

Here is a handy graph by ikawrakow comparing some lower-quality quant
types (lower is better):

image.png

And here are Artefact2's thoughts on the matter:
https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9

FAQ / Model Request

See https://huggingface.co/mradermacher/model_requests for some answers to
questions you might have and/or if you want some other model quantized.

Thanks

I thank my company, nethype GmbH, for letting
me use its servers and providing upgrades to my workstation to enable
this work in my free time. Additional thanks to @nicoboss for giving me access to his private supercomputer, enabling me to provide many more imatrix quants, at much higher quality, than I would otherwise be able to.

README history 3 versions

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

  1. 2026-11-12 auto-patch README.md 2ae7191 6.8 KB
    Loading...
  2. 2026-11-12 auto-patch README.md 23b9449 4.2 KB
    Loading...
  3. 2026-11-12 uploaded from rich1 38bd18b 491 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