Qwen3.8-27B-Humanlike-Chat
27BQwenuncensored
from huihui-ai/Huihui-Qwen3.8-27B-abliterated
↓ 86.4K♥ 2253 repos
Qwen3.8-27B-abliterated-NInfer-v3
Qwen3.8-27B-abliterated-NInfer-v3 is a refusal-ablated variant of huihui-ai/Huihui-Qwen3.8-27B-abliterated, published on Hugging Face by WaveCut. It is 27B parameters, 48GB VRAM class class and apache-2.0 licence. It is distributed across 3 repositories in Transformers and NVFP4 formats, totalling 16.8K downloads.
Specification
- Model name
- Qwen3.8-27B-abliterated-NInfer-v3
- Source repository
- WaveCut/Huihui-Qwen3.8-27B-abliterated-NInfer-v3
- Base model
- huihui-ai/Huihui-Qwen3.8-27B-abliterated
- Publisher
- WaveCut
- Parameters
- 27B
- VRAM required
- ~17.2 GB at Q4_K_M, 4k context (estimated) — runs on 24 GB cards
- Hardware class
- 25B – 50B — 48GB VRAM class
- Licence
- apache-2.0
- Task
- Image text to text
- Context window
- 8K tokens (as documented)
- Ablation scope
- Not stated by the publisher
- Formats available
- Transformers, NVFP4
- First indexed
- 21 Sept 2026
- Last updated
- 6 Oct 2026
Ablation technique
How Qwen3.8-27B-abliterated-NInfer-v3 was modified, and what that implies.
Abliteration (directional ablation)
The single residual-stream direction that mediates refusal is identified from harmful/harmless prompt pairs and projected out of the model weights.
Reported scope: Not stated by the publisher. Publishers frequently omit this, so verify behaviour empirically rather than assuming full coverage.
From the publisher’s model card
“the abliterated Qwen3.8-27B with the lowest reported divergence from the base checkpoint among the widely used releases (the first fifteen layers are untouched; independent KL divergence measurements sit around 0.008) and a working MTP head.”
Running it
Commands are templates — confirm the exact repository and quant file before use.
Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "WaveCut/Huihui-Qwen3.8-27B-abliterated-NInfer-v3"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve WaveCut/Huihui-Qwen3.8-27B-abliterated-NInfer-v3 --trust-remote-codeDownloads and variants (2)
Every published repository of Qwen3.8-27B-abliterated-NInfer-v3, including quantised re-releases by other authors.
| Repository | Format | Downloads |
|---|---|---|
| WaveCut/Huihui-Qwen3.8-27B-abliterated-NInfer-v3source | Transformers | 14.3K |
| satellitedown/Huihui-Qwen3.8-27B-abliterated-NVFP4-NInfer-v3 | NVFP4 | 2.5K |
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- Qwen3.8-27B-abliterated-NInfer-v3 will attempt offensive-security prompts that a hosted commercial model declines, which is what makes it usable for red-team corpus generation and for measuring what an unaligned model of this class produces.
- Capability is inherited from huihui-ai/Huihui-Qwen3.8-27B-abliterated, not added by ablation — at 27B parameters, expect roughly the base model’s competence, minus some instruction-following fidelity.
- The publisher does not state which layers were ablated, so assume nothing about refusal consistency — probe it directly.
- Declared licence is apache-2.0, but the terms that bind you are huihui-ai/Huihui-Qwen3.8-27B-abliterated’s — a re-release cannot grant more than its parent.
Handling: run it isolated, put moderation in front of it if anyone outside your team can reach it, and verify the weights before loading — these are third-party artifacts. Full handling and licence guidance.
Metadata for Qwen3.8-27B-abliterated-NInfer-v3 is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with WaveCut, and does not independently verify publisher claims. See responsible use.