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

noctrex/Llama-3.3-8B-Instruct-128k-abliterated-GGUF

Name: noctrex/Llama-3.3-8B-Instruct-128k-abliterated-GGUF
Author: noctrex

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/noctrex%2FLlama-3.3-8B-Instruct-128k-abliterated-GGUF"

Response includes

classification m8
files 12
benchmarks 11 entries
hub_downloads_all_time 7,838
author_summary 44 models
readme_text full

10 credits · hourly refresh · ~4 KB payload Get an API key →

Abliteration classifier · v1.0.0

Primary method

Repackaging (quantization)

No other method signals detected in this model.

Confidence

HIGH

Why this label 3 signals

Producer identity confirmed by naming conventions, tags or the model card. This label is very unlikely to change.

author=noctrex (M8 quantization producer)
is_gguf=1
base_model='allura-forge/Llama-3.3-8B-Instruct' (source unknown method)

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.

Downloads · lifetime

551 last 30d - cooling

Likes

Model age

9mo ago

created 2025-12-31

Downloads over time

Dec 31, 2025 → Oct 12 · 81 snapshots · spans 285 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	1.7	UGI
Hazardous	1.8	UGI
Natural Intelligence	16.46	UGI
Political lean	-21.1%	UGI
Sensitive-Info	20.83	UGI
SocPol	2.8	UGI
UGI	25.55	UGI
Willingness (10)	3.5	UGI
W10-Adherence	5	UGI
W10-Direct	2	UGI
Writing	9.75	UGI

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.

allura-forge/Llama-3.3-8B-Instruct this lineage

noctrex/Llama-3.3-8B-Instruct-128k-abliterated-GGUF↓ 551

Metadata

Quantizations

BF16 F16 IQ3 IQ4 Q4_K Q5_K Q6_K Q8_0

Files by quantization

BF16 1 file 15.0 GB

Llama-3.3-8B-Instruct-128k-abliterated-BF16.gguf

15.0 GB

af9db43c

download

F16 1 file 15.0 GB

Llama-3.3-8B-Instruct-128k-abliterated-F16.gguf

15.0 GB

a59f38eb

download

Q8_0 1 file 7.95 GB

Llama-3.3-8B-Instruct-128k-abliterated-Q8_0.gguf

7.95 GB

cdb5b4ab

download

Q6_K 1 file 6.14 GB

Llama-3.3-8B-Instruct-128k-abliterated-Q6_K.gguf

6.14 GB

f4eff36b

download

Q5_K 1 file 5.34 GB

Llama-3.3-8B-Instruct-128k-abliterated-Q5_K_M.gguf

5.34 GB

19f248c1

download

Q4_K 1 file 4.58 GB

Llama-3.3-8B-Instruct-128k-abliterated-Q4_K_M.gguf

4.58 GB

8bd94b8c

download

IQ4 2 files 8.50 GB

Llama-3.3-8B-Instruct-128k-abliterated-IQ4_NL.gguf	4.36 GB	a1ab8550	download
Llama-3.3-8B-Instruct-128k-abliterated-IQ4_XS.gguf	4.14 GB	658bde3c	download

IQ3 1 file 3.52 GB

Llama-3.3-8B-Instruct-128k-abliterated-IQ3_M.gguf

3.52 GB

afe592bf

download

Auxiliary files 3 files 4.79 MB

imatrix.gguf	4.78 MB	0364c1ff	download
.gitattributes	2.29 KB	5f78bc01	download
README.md	1.17 KB	df755b0b	download

README current version from Hugging Face

pipeline_tag: text-generation
tags:

uncensored
abliterated
base_model:
allura-forge/Llama-3.3-8B-Instruct

This is an abliterated version of Llama-3.3-8B-Instruct, made using Heretic v1.1.0

This is the model with 128k context.

The quantizations were created using an imatrix merged from combined_en_medium and harmful.txt to leverage the abliterated nature of the model.

Performance

Metric	This model	Original model
Refusals	5/100	95/100
KL divergence	0.0919	0

BibTeX entry and citation info

@misc{heretic,
  author = {Weidmann, Philipp Emanuel},
  title = {Heretic: Fully automatic censorship removal for language models},
  year = {2025},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/p-e-w/heretic}}
}

README history 2 versions

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

2025-12-31Add files using upload-large-folder toola2fb54b1.2 KB
Loading...
2025-12-31Create README.mde75db051.2 KB
Loading...

noctrex/Llama-3.3-8B-Instruct-128k-abliterated-GGUF

The browser can't see your hardware

The Abliteration app is integrated with your machine

And you can chat with any model, right now

Repackaging (quantization)

Benchmarks

Genealogy 0 direct forks

Metadata

Related

Files by quantization

README current version from Hugging Face

Performance

BibTeX entry and citation info

README history 2 versions

Run models on your own machine, not in the cloud.