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

nitrox/Qwen3-4B-abliterated-v2

nitrox Qwen 4.0B MoE
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/nitrox%2FQwen3-4B-abliterated-v2"
Response includes
  • classification m1
  • files 14
  • hub_downloads_all_time 64
  • author_summary 3 models
  • readme_text full
Try next
GET/models?author=nitrox
All 3 models by nitrox
 GET/models?family=qwen&method=m1
Sibling qwen 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
64
15 last 30d - stable
Likes
0
Descendants
2
in 2 direct forks
Model age
7mo ago
created 2026-02-20
Downloads over time
Now68→from9↑656%
62951749 on Feb 1868 on Oct 1168 on Oct 9FebAprJunAugOct
Feb 18 → Oct 11 · 73 snapshots · spans 235 days

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.

INC4AI/MiniMax-M2.5-int4-mixed-AutoRound this lineage
nitrox/Qwen3-4B-abliterated-v2↓ 15 2 forks
mradermacher/Qwen3-4B-abliterated-v2-i1-GGUF↓ 1,064
mradermacher/Qwen3-4B-abliterated-v2-GGUF↓ 619

Metadata

License
apache-2.0
Tags
safetensors qwen3 abliteration uncensored minimax moe int4 base_model:Intel/MiniMax-M2.5-int4-AutoRound base_model:finetune:Intel/MiniMax-M2.5-int4-AutoRound license:apache-2.0 region:us

Related

Other by nitrox
MiniMax-M2.5-abliterated-Q4_K_M-GGUF 20 SA-SWE-32B-abliterated 9
Total size
7.49 GB
Files
14
Quantizations
1
Registered
2026-08-22 13:56
Last updated on HF
2026-02-20 22:29
Source huggingface.co/nitrox/Qwen3-4B-abliterated-v2 ↗ Base model INC4AI/MiniMax-M2.5-int4-mixed-AutoRound ↗

Files by quantization

Auxiliary files 14 files 7.51 GB
model-00001-of-00002.safetensors 4.63 GB afe46f6c download
model-00002-of-00002.safetensors 2.87 GB 11f912fc download
tokenizer.json 10.9 MB 67cc0080 download
vocab.json 2.65 MB 4783fe10 download
merges.txt 1.59 MB 31349551 download
model.safetensors.index.json 32.1 KB b65d8063 download
tokenizer_config.json 5.28 KB c9fc1221 download
chat_template.jinja 2.57 KB 70adff8a download
.gitattributes 1.53 KB 52373fe2 download
config.json 1.50 KB 316f85a6 download
README.md 1.12 KB cb5d075d download
added_tokens.json 707 B b54f9135 download
special_tokens_map.json 613 B ac23c0aa download
generation_config.json 213 B 3dd2c545 download

README current version from Hugging Face

license: apache-2.0
tags:

  • abliteration
  • uncensored
  • minimax
  • moe
  • int4
    base_model: INC4AI/MiniMax-M2.5-int4-mixed-AutoRound

MiniMax-M2.5 Abliterated (int4)

This is an abliterated version of INC4AI/MiniMax-M2.5-int4-mixed-AutoRound.

Abliteration

Abliteration was performed using heretic — a multi-objective optimization framework that uses Optuna TPE to find the best LoRA-based abliteration parameters.

  • Method: Heretic v1.2.0, LoRA + Optuna multi-objective optimization
  • Base model: INC4AI/MiniMax-M2.5-int4-mixed-AutoRound (230B MoE)
  • Format: int4 AutoRound (Marlin backend)

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained(
    "nitrox/Qwen3-4B-abliterated-v2",
    device_map="auto",
    trust_remote_code=True,
)
tokenizer = AutoTokenizer.from_pretrained("nitrox/Qwen3-4B-abliterated-v2", trust_remote_code=True)

Disclaimer

This model has had its refusal mechanisms removed. Use responsibly.

README history 2 versions

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

  1. 2026-02-20Upload README.md with huggingface_hub74df5021.1 KB
    Loading...
  2. 2026-02-20initial commit362040528 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