Qwen3.5-9B-The-Defiant-Fable-Uncensored-Heretic
9BQwenheretic
from nightmedia/Qwen3.5-9B-DS9-USS-Defiant
↓ 2M♥ 1.4K29 repos
Heretical-Qwen3.5-9B-convrot-Comfy
Heretical-Qwen3.5-9B-convrot-Comfy is a Heretic-ablated variant of Kewk/Heretical-Qwen3.5-9B, published on Hugging Face by rzgar. It is 9B parameters, single consumer GPU class and apache-2.0 licence. It is distributed across 2 repositories in Transformers format, totalling 5.4K downloads.
Specification
- Model name
- Heretical-Qwen3.5-9B-convrot-Comfy
- Source repository
- rzgar/Heretical-Qwen3.5-9B-int8-convrot-Comfy
- Base model
- Kewk/Heretical-Qwen3.5-9B
- Publisher
- rzgar
- 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
- Image text to text
- Context window
- Not documented
- Ablation scope
- Not stated by the publisher
- Formats available
- Transformers
- First indexed
- 11 Oct 2026
- Last updated
- 11 Oct 2026
Ablation technique
How Heretical-Qwen3.5-9B-convrot-Comfy 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
“Weight error vs the bf16 source (relative L2, over all 200 quantized layers): 0.0090”
Running it
Commands are templates — confirm the exact repository and quant file before use.
Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "rzgar/Heretical-Qwen3.5-9B-int8-convrot-Comfy"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve rzgar/Heretical-Qwen3.5-9B-int8-convrot-Comfy --trust-remote-codeDownloads and variants (1)
The only published repository of Heretical-Qwen3.5-9B-convrot-Comfy.
| Repository | Format | Downloads |
|---|---|---|
| rzgar/Heretical-Qwen3.5-9B-int8-convrot-Comfysource | Transformers | 5.4K |
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- Heretical-Qwen3.5-9B-convrot-Comfy 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 Kewk/Heretical-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 Kewk/Heretical-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 Heretical-Qwen3.5-9B-convrot-Comfy is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with rzgar, and does not independently verify publisher claims. See responsible use.