Qwen3-VL-8B-Instruct-c_abliterated-v3
8BQwengated
from prithivMLmods/Qwen3-VL-8B-Instruct-abliterated-v2
↓ 9.4K♥ 96 repos
Qwen3-VL-8B-Instruct-abliterated-v2.0
Qwen3-VL-8B-Instruct-abliterated-v2.0 is a refusal-ablated variant of prithivMLmods/Qwen3-VL-8B-Instruct-abliterated-v2, published on Hugging Face by mradermacher. It is 8B parameters, single consumer GPU class and apache-2.0 licence. It is distributed across 3 repositories in GGUF and GGUF (imatrix) formats, totalling 11.1K downloads.
Specification
- Model name
- Qwen3-VL-8B-Instruct-abliterated-v2.0
- Source repository
- mradermacher/Qwen3-VL-8B-Instruct-abliterated-v2.0-GGUF
- Publisher
- mradermacher
- Parameters
- 8B
- Hardware class
- 4B – 10B — single consumer GPU
- Licence
- apache-2.0
- Task
- Unspecified
- Context window
- Not documented
- Ablation scope
- Not stated by the publisher
- Formats available
- GGUF, GGUF (imatrix)
- First indexed
- 10 Nov 2025
- Last updated
- 4 Dec 2025
Ablation technique
How Qwen3-VL-8B-Instruct-abliterated-v2.0 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
“For a convenient overview and download list, visit our model page for this model.”
Running it
Commands are templates — confirm the exact repository and quant file before use.
llama.cpp / GGUF
hf download mradermacher/Qwen3-VL-8B-Instruct-abliterated-v2.0-GGUF --local-dir ./qwen3-vl-8b-instruct-abliterated-v2.0
llama-cli -m ./qwen3-vl-8b-instruct-abliterated-v2.0/<file>.gguf -p "..."Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "mradermacher/Qwen3-VL-8B-Instruct-abliterated-v2.0-GGUF"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve mradermacher/Qwen3-VL-8B-Instruct-abliterated-v2.0-GGUF --trust-remote-codeDownloads and variants (2)
Every published repository of Qwen3-VL-8B-Instruct-abliterated-v2.0, including quantised re-releases by other authors.
| Repository | Format | Downloads |
|---|---|---|
| mradermacher/Qwen3-VL-8B-Instruct-abliterated-v2.0-GGUFsource | GGUF | 9.9K |
| mradermacher/Qwen3-VL-8B-Instruct-abliterated-v2.0-i1-GGUF | GGUF (imatrix) | 1.2K |
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- Qwen3-VL-8B-Instruct-abliterated-v2.0 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 prithivMLmods/Qwen3-VL-8B-Instruct-abliterated-v2, not added by ablation — at 8B parameters, expect a small model’s reasoning and a high error rate on exploit detail.
- 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 prithivMLmods/Qwen3-VL-8B-Instruct-abliterated-v2’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-VL-8B-Instruct-abliterated-v2.0 is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with mradermacher, and does not independently verify publisher claims. See responsible use.