⌦abliteratedmodels.org

Qwen3.8-27B-Fable-Distill-Heretic-ara

Qwen3.8-27B-Fable-Distill-Heretic-ara is a Heretic-ablated variant of TeichAI/Qwen3.8-27B-Fable-Distill, published on Hugging Face by armand0e. It is 27B parameters and 48GB VRAM class class. It is distributed across 6 repositories in Transformers, GGUF and GGUF (imatrix) formats, totalling 15.8K downloads.

Heretic27BQwen25B – 50B↓ 15,813♥ 7
View on Hugging Face ↗Base model ↗

Specification

Model name
Qwen3.8-27B-Fable-Distill-Heretic-ara
Source repository
armand0e/Qwen3.8-27B-Fable-Distill-Heretic-ara
Base model
TeichAI/Qwen3.8-27B-Fable-Distill
Publisher
armand0e
Parameters
27B
Hardware class
25B – 50B — 48GB VRAM class
Licence
Not declared
Task
Image text to text
Context window
Not documented
Ablation scope
Not stated by the publisher
Formats available
Transformers, GGUF, GGUF (imatrix)
First indexed
16 Aug 2026
Last updated
23 Aug 2026

Ablation technique

How Qwen3.8-27B-Fable-Distill-Heretic-ara 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

“The ablation is a rank-2 update applied to the attn.oproj, mlp.downproj projections of layers 30-64, solved in closed form rather than by gradient descent.”

Running it

Commands are templates — confirm the exact repository and quant file before use.

llama.cpp / GGUF
hf download armand0e/Qwen3.8-27B-Fable-Distill-Heretic-ara-GGUF --local-dir ./qwen3.8-27b-fable-distill-heretic-ara
llama-cli -m ./qwen3.8-27b-fable-distill-heretic-ara/<file>.gguf -p "..."
Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer

repo = "armand0e/Qwen3.8-27B-Fable-Distill-Heretic-ara"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
    repo, torch_dtype="auto", device_map="auto"
)
vLLM
vllm serve armand0e/Qwen3.8-27B-Fable-Distill-Heretic-ara --trust-remote-code

Downloads and variants (5)

Every published repository of Qwen3.8-27B-Fable-Distill-Heretic-ara, including quantised re-releases by other authors.

RepositoryFormatDownloads
armand0e/Qwen3.8-27B-Fable-Distill-Heretic-ara-GGUFGGUF7.3K
mradermacher/Qwen3.8-27B-Fable-Distill-Heretic-ara-i1-GGUFGGUF (imatrix)5.4K
mradermacher/Qwen3.8-27B-Fable-Distill-Heretic-ara-GGUFGGUF2.3K
brindusa/Qwen3.8-27B-Fable-Distill-Heretic-ara-GGUFGGUF430
armand0e/Qwen3.8-27B-Fable-Distill-Heretic-arasourceTransformers375

Security-team assessment

Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.

  • Qwen3.8-27B-Fable-Distill-Heretic-ara 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 TeichAI/Qwen3.8-27B-Fable-Distill, not added by ablation — at 27B parameters, expect roughly the base model’s competence, minus some instruction-following fidelity.
  • The publisher does not state which layers were ablated, so assume nothing about refusal consistency — probe it directly.
  • No licence is declared on this repository, which is a compliance problem in itself. Resolve it against the base model’s terms before use.

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 Qwen3.8-27B-Fable-Distill-Heretic-ara is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with armand0e, and does not independently verify publisher claims. See responsible use.