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

darkmaniac7/Qwen3.5-4B-Claude-4.6-Opus-abliterated-MNN

Name: darkmaniac7/Qwen3.5-4B-Claude-4.6-Opus-abliterated-MNN
Author: darkmaniac7

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/darkmaniac7%2FQwen3.5-4B-Claude-4.6-Opus-abliterated-MNN"

Response includes

classification m1
files 9
benchmarks 11 entries
hub_downloads_all_time 441
author_summary 21 models
readme_text full

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

Abliteration classifier · v1.0.0

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)

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)

Downloads · lifetime

441

78 last 30d - stable

Likes

Model age

6mo ago

created 2026-03-20

Downloads over time

Mar 18 → Oct 11 · 69 snapshots · spans 207 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	0.9	UGI
Hazardous	1.2	UGI
Natural Intelligence	13.45	UGI
Political lean	-17.3%	UGI
Sensitive-Info	11.73	UGI
SocPol	1.5	UGI
UGI	15.32	UGI
Willingness (10)	2.2	UGI
W10-Adherence	1.5	UGI
W10-Direct	3	UGI
Writing	29.68	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.

Qwen/Qwen3.5-4B this lineage

darkmaniac7/Qwen3.5-4B-Claude-4.6-Opus-abliterated-MNN↓ 78

Metadata

License

apache-2.0

Files by quantization

Auxiliary files 9 files 2.22 GB

llm.mnn.weight	2.20 GB	0b206c42	download
llm.mnn.json	8.75 MB	f33cd741	download
llm.mnn	3.48 MB	9d6ed852	download
tokenizer.txt	2.82 MB	a7d4b348	download
llm_config.json	4.43 KB	2aab8545	download
README.md	1.88 KB	5346ce1e	download
.gitattributes	1.58 KB	e547efe9	download
export_args.json	1.14 KB	24e984a3	download
config.json	210 B	bd685f44	download

README current version from Hugging Face

license: apache-2.0
tags:

mnn
qwen3.5
mobile
on-device
tokforge
abliterated
linear-attention
base_model: Qwen/Qwen3.5-4B

TokForge

Website: https://tokforge.ai
Discord: https://discord.gg/Acv3CBtfVm
Google Play: https://play.google.com/store/apps/details?id=dev.tokforge
iOS TestFlight: https://testflight.apple.com/join/jnufjzRr

Runs on-device in the TokForge app.

Qwen3.5-4B-Claude-4.6-Opus-abliterated (MNN)

Pre-converted Qwen3.5-4B abliterated model in MNN format for on-device inference.

Model Details

Architecture: Qwen3.5 (hybrid LinearAttention + standard attention)
Parameters: 4B (4-bit quantized)
Format: MNN (Alibaba Mobile Neural Network)
Quantization: W4A16 (4-bit weights, 16-bit activations)
Attention: LinearAttention (gated_delta_rule) + standard attention hybrid

Performance

Device	SoC	Backend	tok/s
S26 Ultra	SM8850	CPU	20.5
Lenovo TB520FU	SM8650	CPU	14.0
Xiaomi Pad 7 Pro	SM8635	CPU	11.8

CPU is recommended for Qwen3.5 models (LinearAttention runs natively on CPU).

Usage with TokForge

Optimized for TokForge — an Android app for on-device LLM inference.

Abliteration

Safety filters removed for unrestricted conversation. Use responsibly.

Limitations and Intended Use

Intended for TokForge / MNN on-device inference, especially creative writing and roleplay use.
Qwen3.5 hybrid LinearAttention models generally route best to CPU in current TokForge builds.
Performance varies by device class and generation length.
This repo is a packaged runtime artifact, not a standard Transformers training checkpoint.

Community

Website: tokforge.ai
Discord: Join the Discord

README history 4 versions

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

2026-10-05Card metadata: license, base_model, base_model_relation323a7eb1.9 KB
Loading...
2026-07-04Add TokForge app linksd51628b1.9 KB
Loading...
2026-03-25Upload README.md with huggingface_hubb4b20cd1.6 KB
Loading...
2026-03-20Upload folder using huggingface_hube84377a1.1 KB
Loading...