⌦abliteratedmodels.org

Stellar-Witch-12B-Heretic

Stellar-Witch-12B-Heretic is a Heretic-ablated variant of Vortex5/Stellar-Witch-12B, published on Hugging Face by Sorihon. It is 12B parameters and 24GB VRAM class class. It is distributed across 4 repositories in Transformers, GGUF and GGUF (imatrix) formats, totalling 7.9K downloads.

Heretic12BOther10B – 25B↓ 7,923♥ 6
View on Hugging Face ↗Base model ↗

Specification

Model name
Stellar-Witch-12B-Heretic
Source repository
Sorihon/Stellar-Witch-12B-Heretic
Base model
Vortex5/Stellar-Witch-12B
Publisher
Sorihon
Parameters
12B
VRAM required
~8.1 GB at Q4_K_M, 4k context (estimated) — runs on 12 GB cards
Hardware class
10B – 25B — 24GB VRAM class
Licence
Not declared
Task
Unspecified
Context window
Not documented
Ablation scope
Not stated by the publisher
Formats available
Transformers, GGUF, GGUF (imatrix)
First indexed
28 Apr 2026
Last updated
29 Apr 2026

Ablation technique

How Stellar-Witch-12B-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.

Running it

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

llama.cpp / GGUF
hf download mradermacher/Stellar-Witch-12B-Heretic-GGUF --local-dir ./stellar-witch-12b-heretic
llama-cli -m ./stellar-witch-12b-heretic/<file>.gguf -p "..."
Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer

repo = "Sorihon/Stellar-Witch-12B-Heretic"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
    repo, torch_dtype="auto", device_map="auto"
)
vLLM
vllm serve Sorihon/Stellar-Witch-12B-Heretic --trust-remote-code

Downloads and variants (3)

Every published repository of Stellar-Witch-12B-Heretic, including quantised re-releases by other authors.

RepositoryFormatDownloads
mradermacher/Stellar-Witch-12B-Heretic-GGUFGGUF5.5K
mradermacher/Stellar-Witch-12B-Heretic-i1-GGUFGGUF (imatrix)2.4K
Sorihon/Stellar-Witch-12B-HereticsourceTransformers16

Security-team assessment

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

  • Stellar-Witch-12B-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 Vortex5/Stellar-Witch-12B, not added by ablation — at 12B 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 Stellar-Witch-12B-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 Sorihon, and does not independently verify publisher claims. See responsible use.