Qwen3.5-9B-The-Defiant-Fable-Uncensored-Heretic
9BQwenheretic
from nightmedia/Qwen3.5-9B-DS9-USS-Defiant
↓ 2M♥ 1.4K29 repos
InterimLabs-Huihui-Qwen3.5-9B-abliterated
InterimLabs-Huihui-Qwen3.5-9B-abliterated is a refusal-ablated variant of huihui-ai/Huihui-Qwen3.5-9B-abliterated, published on Hugging Face by interimlabs. It is 9B parameters, single consumer GPU class and apache-2.0 licence. It is distributed across 3 repositories in GGUF and MLX formats, totalling 5.3K downloads.
Specification
- Model name
- InterimLabs-Huihui-Qwen3.5-9B-abliterated
- Source repository
- interimlabs/InterimLabs-Huihui-Qwen3.5-9B-abliterated-GGUF-Q4_K_M
- Base model
- huihui-ai/Huihui-Qwen3.5-9B-abliterated
- Publisher
- interimlabs
- Parameters
- 9B
- VRAM required
- ~6.2 GB at Q4_K_M, 4k context (estimated) — runs on 8 GB cards
- Hardware class
- 4B – 10B — single consumer GPU
- Licence
- apache-2.0
- Task
- Text generation
- Context window
- Not documented
- Ablation scope
- Not stated by the publisher
- Formats available
- GGUF, MLX
- First indexed
- 1 Oct 2026
- Last updated
- 8 Oct 2026
Ablation technique
How InterimLabs-Huihui-Qwen3.5-9B-abliterated 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
“A standalone GGUF Q4KM release published by Interim Labs for local text generation. This package derives from huihui-ai/Huihui-Qwen3.5-9B-abliterated. The GGUF weights are redistributed byte-for-byte from the credited conversion source below.”
Running it
Commands are templates — confirm the exact repository and quant file before use.
llama.cpp / GGUF
hf download interimlabs/InterimLabs-Huihui-Qwen3.5-9B-abliterated-GGUF-Q4_K_M --local-dir ./interimlabs-huihui-qwen3.5-9b-abliterated
llama-cli -m ./interimlabs-huihui-qwen3.5-9b-abliterated/<file>.gguf -p "..."Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "interimlabs/InterimLabs-Huihui-Qwen3.5-9B-abliterated-GGUF-Q4_K_M"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve interimlabs/InterimLabs-Huihui-Qwen3.5-9B-abliterated-GGUF-Q4_K_M --trust-remote-codeDownloads and variants (2)
Every published repository of InterimLabs-Huihui-Qwen3.5-9B-abliterated, including quantised re-releases by other authors.
| Repository | Format | Downloads |
|---|---|---|
| interimlabs/InterimLabs-Huihui-Qwen3.5-9B-abliterated-GGUF-Q4_K_Msource | GGUF | 4.6K |
| interimlabs/InterimLabs-Huihui-Qwen3.5-9B-abliterated-MLX-4bit | MLX | 696 |
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- InterimLabs-Huihui-Qwen3.5-9B-abliterated 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-Qwen3.5-9B-abliterated, not added by ablation — at 9B 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 huihui-ai/Huihui-Qwen3.5-9B-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 InterimLabs-Huihui-Qwen3.5-9B-abliterated is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with interimlabs, and does not independently verify publisher claims. See responsible use.