Qwen3.5-9B-Uncensored-HauhauCS-Aggressive
9BQwenuncensored
from Qwen/Qwen3.5-9B
↓ 888.2K♥ 2K83 repos
Qwen3.5-9B-heretic
Qwen3.5-9B-heretic is a Heretic-ablated variant of Qwen/Qwen3.5-9B, published on Hugging Face by coder3101. It is 9B parameters, single consumer GPU class and apache-2.0 licence. It is distributed across 17 repositories in GGUF, Transformers, MLX and GGUF (imatrix) formats, totalling 10.2K downloads.
Specification
- Model name
- Qwen3.5-9B-heretic
- Source repository
- coder3101/Qwen3.5-9B-heretic
- Base model
- Qwen/Qwen3.5-9B
- Publisher
- coder3101
- Parameters
- 9B
- Hardware class
- 4B – 10B — single consumer GPU
- Licence
- apache-2.0
- Task
- Image text to text
- Context window
- Not documented
- Ablation scope
- Not stated by the publisher
- Formats available
- GGUF, Transformers, MLX, GGUF (imatrix), AWQ, EXL3
- First indexed
- 2 Mar 2026
- Last updated
- 5 Sept 2026
Ablation technique
How Qwen3.5-9B-heretic was modified, and what that implies.
Heretic
Automated directional ablation via the Heretic toolchain, which searches for the refusal direction and applies it with a KL-divergence budget so general capability is preserved.
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
“Over recent months, we have intensified our focus on developing foundation models that deliver exceptional utility and performance. Qwen3.5 represents a significant leap forward, integrating breakthroughs in multimodal learning, architectural efficiency, reinforcement learning scale, and global accessibility to...”
Running it
Commands are templates — confirm the exact repository and quant file before use.
llama.cpp / GGUF
hf download mradermacher/Qwen3.5-9B-heretic-GGUF --local-dir ./qwen3.5-9b-heretic
llama-cli -m ./qwen3.5-9b-heretic/<file>.gguf -p "..."Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "coder3101/Qwen3.5-9B-heretic"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve coder3101/Qwen3.5-9B-heretic --trust-remote-codeDownloads and variants (16)
Every published repository of Qwen3.5-9B-heretic, including quantised re-releases by other authors.
| Repository | Format | Downloads |
|---|---|---|
| mradermacher/Qwen3.5-9B-heretic-GGUF | GGUF | 2.8K |
| saidutta69/Qwen3.5-9B-heretic | GGUF | 2.7K |
| mradermacher/Qwen3.5-9B-heretic-i1-GGUF | GGUF (imatrix) | 1.8K |
| FadedRedStar/Qwen3.5-9B-heretic-GGUF | GGUF | 1.4K |
| FadedRedStar/Qwen3.5-9B-heretic-imatrix-GGUF | GGUF | 580 |
| TheCluster/Qwen3.5-9B-Heretic-MLX-mxfp4 | MLX | 439 |
| coder3101/Qwen3.5-9B-hereticsource | Transformers | 324 |
| lemb0t/Qwen3.5-9B-heretic-AWQ-W4A16 | AWQ | 57 |
| MercuriusDream/Qwen3.5-9B-heretic-MLX-nvfp4 | MLX | 24 |
| trohrbaugh/Qwen3.5-9B-heretic | Transformers | 13 |
| TheCluster/Qwen3.5-9B-Heretic-MLX-bf16 | MLX | 11 |
| SassyDiffusion/Qwen3.5-9B-heretic | Transformers | 9 |
| AISafety-Student/Qwen3.5-9B-heretic | Transformers | 3 |
| skilledu/Qwen3.5-9B-heretic | Transformers | 2 |
| dizarian/Qwen3.5-9B-HERETIC-EXL3-2.5bpw | EXL3 | 1 |
| dizarian/Qwen3.5-9B-HERETIC-EXL3-2bpw | EXL3 | 1 |
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- Qwen3.5-9B-heretic 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/Qwen3.5-9B, 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 Qwen/Qwen3.5-9B’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 Qwen3.5-9B-heretic is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with coder3101, and does not independently verify publisher claims. See responsible use.