⌦abliteratedmodels.org

p-e-w_Qwen3-4B-Instruct-2507-heretic

p-e-w_Qwen3-4B-Instruct-2507-heretic is a Heretic-ablated variant of p-e-w/Qwen3-4B-Instruct-2507-heretic, published on Hugging Face by bartowski. It is 4B parameters, single consumer GPU class and apache-2.0 licence. It is distributed across 2 repositories in GGUF format, totalling 8.6K downloads.

Heretic4BOther4B – 10B↓ 8,643♥ 12
View on Hugging Face ↗Base model ↗

Specification

Model name
p-e-w_Qwen3-4B-Instruct-2507-heretic
Source repository
bartowski/p-e-w_Qwen3-4B-Instruct-2507-heretic-GGUF
Base model
p-e-w/Qwen3-4B-Instruct-2507-heretic
Publisher
bartowski
Parameters
4B
VRAM required
~3.2 GB at Q4_K_M, 4k context (estimated) — runs on 4 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
GGUF
First indexed
17 Nov 2025
Last updated
17 Nov 2025

Ablation technique

How p-e-w_Qwen3-4B-Instruct-2507-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

“All quants made using imatrix option with dataset from here combined with a subset of combinedallsmall.parquet from Ed Addario here”

Running it

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

llama.cpp / GGUF
hf download bartowski/p-e-w_Qwen3-4B-Instruct-2507-heretic-GGUF --local-dir ./p-e-w-qwen3-4b-instruct-2507-heretic
llama-cli -m ./p-e-w-qwen3-4b-instruct-2507-heretic/<file>.gguf -p "..."
Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer

repo = "bartowski/p-e-w_Qwen3-4B-Instruct-2507-heretic-GGUF"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
    repo, torch_dtype="auto", device_map="auto"
)
vLLM
vllm serve bartowski/p-e-w_Qwen3-4B-Instruct-2507-heretic-GGUF --trust-remote-code

Downloads and variants (1)

The only published repository of p-e-w_Qwen3-4B-Instruct-2507-heretic.

RepositoryFormatDownloads
bartowski/p-e-w_Qwen3-4B-Instruct-2507-heretic-GGUFsourceGGUF8.6K

Security-team assessment

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

  • p-e-w_Qwen3-4B-Instruct-2507-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 p-e-w/Qwen3-4B-Instruct-2507-heretic, not added by ablation — at 4B 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 p-e-w/Qwen3-4B-Instruct-2507-heretic’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 p-e-w_Qwen3-4B-Instruct-2507-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 bartowski, and does not independently verify publisher claims. See responsible use.