Qwen3-14B-abliterated
14BQwen
from Qwen/Qwen3-14B
↓ 69.9K♥ 5620 repos
Qwen3-14B-abliterated-v2
Qwen3-14B-abliterated-v2 is a refusal-ablated variant of Qwen/Qwen3-14B, published on Hugging Face by huihui-ai. It is 14B parameters, 24GB VRAM class class and apache-2.0 licence. It is distributed across 10 repositories in GGUF, Transformers, GGUF (imatrix) and AWQ formats, totalling 61.9K downloads.
Specification
- Model name
- Qwen3-14B-abliterated-v2
- Source repository
- huihui-ai/Huihui-Qwen3-14B-abliterated-v2
- Base model
- Qwen/Qwen3-14B
- Publisher
- huihui-ai
- Parameters
- 14B
- VRAM required
- ~9.3 GB at Q4_K_M, 4k context (estimated) — runs on 12 GB cards
- Hardware class
- 10B – 25B — 24GB VRAM class
- Licence
- apache-2.0
- Task
- Text generation
- Context window
- Not documented
- Ablation scope
- Not stated by the publisher
- Formats available
- GGUF, Transformers, GGUF (imatrix), AWQ, EXL2, MLX
- First indexed
- 17 Jun 2025
- Last updated
- 15 Sept 2026
Ablation technique
How Qwen3-14B-abliterated-v2 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
“This is an uncensored version of Qwen/Qwen3-14B created with abliteration (see remove-refusals-with-transformers to know more about it). This is a crude, proof-of-concept implementation to remove refusals from an LLM model without using TransformerLens.”
Running it
Commands are templates — confirm the exact repository and quant file before use.
Ollama (publisher-documented)
ollama run huihui_ai/qwen3-abliterated:14b-v2llama.cpp / GGUF
hf download mradermacher/Huihui-Qwen3-14B-abliterated-v2-GGUF --local-dir ./qwen3-14b-abliterated-v2
llama-cli -m ./qwen3-14b-abliterated-v2/<file>.gguf -p "..."Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "huihui-ai/Huihui-Qwen3-14B-abliterated-v2"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve huihui-ai/Huihui-Qwen3-14B-abliterated-v2 --trust-remote-codeDownloads and variants (9)
Every published repository of Qwen3-14B-abliterated-v2, including quantised re-releases by other authors.
| Repository | Format | Downloads |
|---|---|---|
| mradermacher/Huihui-Qwen3-14B-abliterated-v2-GGUF | GGUF | 50.4K |
| mradermacher/Huihui-Qwen3-14B-abliterated-v2-i1-GGUF | GGUF (imatrix) | 7.2K |
| huihui-ai/Huihui-Qwen3-14B-abliterated-v2source | Transformers | 2.7K |
| warshanks/Huihui-Qwen3-14B-abliterated-v2-AWQ | AWQ | 1K |
| delpheye/Huihui-Qwen3-14B-abliterated-v2-Q4_K_M | GGUF | 285 |
| pj1983/Huihui-Qwen3-14B-abliterated-v2-mlx-4bit | MLX | 178 |
| pj1983/Huihui-Qwen3-14B-abliterated-v2-mlx-5bit | MLX | 84 |
| cgus/Huihui-Qwen3-14B-abliterated-v2-exl2 | EXL2 | 66 |
| harimuliawan999/Huihui-Qwen3-14B-abliterated-v2 | Transformers | 33 |
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- Qwen3-14B-abliterated-v2 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 Qwen/Qwen3-14B, not added by ablation — at 14B 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 Qwen/Qwen3-14B’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 Qwen3-14B-abliterated-v2 is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with huihui-ai, and does not independently verify publisher claims. See responsible use.