Qwen3.5-9B-The-Defiant-Fable-Uncensored-Heretic
9BQwenheretic
from nightmedia/Qwen3.5-9B-DS9-USS-Defiant
↓ 2M♥ 1.4K29 repos
qwen3-8b-apostate
qwen3-8b-apostate is a abliterated and uncensored variant of heterodoxin/qwen3-8b-apostate, published on Hugging Face by heterodoxin. It is 8B parameters, single consumer GPU class and apache-2.0 licence. It is distributed across 5 repositories in Transformers, GGUF and GGUF (imatrix) formats, totalling 6.9K downloads.
Specification
- Model name
- qwen3-8b-apostate
- Source repository
- heterodoxin/qwen3-8b-apostate
- Base model
- heterodoxin/qwen3-8b-apostate
- Publisher
- heterodoxin
- Parameters
- 8B
- VRAM required
- ~5.6 GB at Q4_K_M, 4k context (estimated) — runs on 6 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
- Transformers, GGUF, GGUF (imatrix)
- First indexed
- 22 Jun 2026
- Last updated
- 24 Jun 2026
Ablation technique
How qwen3-8b-apostate was modified, and what that implies.
Abliteration + uncensored fine-tune
Directional ablation combined with additional fine-tuning on unfiltered data, so behaviour diverges from the base model beyond refusal removal alone.
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
“An uncensored edit of Qwen/Qwen3-8B. Refusal behavior is removed by editing the weights directly — no finetuning, no adapter, no runtime hook. The result is a standard Transformers checkpoint that loads anywhere Qwen3 does.”
Running it
Commands are templates — confirm the exact repository and quant file before use.
llama.cpp / GGUF
hf download mradermacher/qwen3-8b-apostate-GGUF --local-dir ./qwen3-8b-apostate
llama-cli -m ./qwen3-8b-apostate/<file>.gguf -p "..."Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "heterodoxin/qwen3-8b-apostate"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve heterodoxin/qwen3-8b-apostate --trust-remote-codeDownloads and variants (4)
Every published repository of qwen3-8b-apostate, including quantised re-releases by other authors.
| Repository | Format | Downloads |
|---|---|---|
| mradermacher/qwen3-8b-apostate-GGUF | GGUF | 6.1K |
| mradermacher/qwen3-8b-apostate-i1-GGUF | GGUF (imatrix) | 562 |
| heterodoxin/qwen3-8b-apostate-gguf | GGUF | 263 |
| heterodoxin/qwen3-8b-apostatesource | Transformers | 13 |
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- qwen3-8b-apostate 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 heterodoxin/qwen3-8b-apostate, 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.
- This release adds fine-tuning on unfiltered data, so its behaviour diverges from the base model beyond refusal removal — differences you measure cannot be attributed to ablation alone.
- Declared licence is apache-2.0, but the terms that bind you are heterodoxin/qwen3-8b-apostate’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 qwen3-8b-apostate is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with heterodoxin, and does not independently verify publisher claims. See responsible use.