← back to catalog · registered 2026-09-12 09:55

Bojun-Feng/Qwen3.8-27B-Heretic-ARA-Uncensored-llamafile

Bojun-Feng 27B GGUF second-order
Abliteration classifier · v1.0.0
M3
Primary method

Layer-wise ablation

Applied on top of direct removal inherited from the base model.
Confidence
HIGH
Inherited from base model
Why this label 2 signals
Producer identity confirmed by naming conventions, tags or the model card. This label is very unlikely to change.
  • 'heretic' in model name (Heretic-produced)
  • Heretic uses layer-wise optimization (M3) with underlying direction removal (M1)
Refusal direction extracted via
Extraction technique

Difference-of-means

Confidence
HIGH
Why we say so
name contains 'heretic'; Heretic default extraction is difference-of-means (Arditi 2024)
Downloads · 30-day
107
Likes
0
Model age
2d ago
created 2026-09-10

Genealogy 0 direct forks

Full fork graph →

This model's place in the market. Above: what it was derived from. Below: the tree of everything derived from it.

Metadata

License
apache-2.0
Languages
en
Tags
llamafile gguf qwen3.8 text-generation en base_model:trohrbaugh/Qwen3.8-27B-heretic-ara base_model:finetune:trohrbaugh/Qwen3.8-27B-heretic-ara license:apache-2.0 region:us

Related

Total size
0 B
Files
13
Quantizations
1
Registered
2026-09-12 09:55
Last updated on HF
2026-09-12 10:20

Files by quantization

Auxiliary files 13 files 96.0 GB
Qwen3.8-27B-heretic-ara.Q8_0.llamafile 27.4 GB 988fde19 download
Qwen3.8-27B-heretic-ara.Q6_K.llamafile 21.2 GB 08b43111 download
Qwen3.8-27B-heretic-ara.Q5_K_M.llamafile 18.5 GB 31a653a6 download
Qwen3.8-27B-heretic-ara.Q4_K_M.llamafile 16.0 GB 3d5212c7 download
Qwen3.8-27B-heretic-ara.Q3_K_M.llamafile 12.9 GB f5eceb17 download
LICENSE 11.3 KB f938136e download
manifest.json 7.20 KB bb601164 download
README.md 3.59 KB dce069fe download
PROVENANCE.json 1.75 KB ce776070 download
LICENSE.llamafile 583 B 7661d5fd download
SHA256SUMS 531 B 983f5d7d download
BUILDINFO.json 328 B 13c83a05 download
.gitattributes 48.0 B 9da5eb2f download

README current version from Hugging Face


license: apache-2.0
pipeline_tag: text-generation
base_model: trohrbaugh/Qwen3.8-27B-heretic-ara
tags:

  • llamafile
  • gguf
  • qwen3.8
    language:
  • en

LlamaFile

I am not the original creator of llamafile, all credit of llamafile goes to Jartine:

jartine's LLM work is generously supported by a grant from mozilla


Trohrbaugh Heretic ARA GGUF - llamafile

Run LLMs locally with a single file

Download a file and run it. Llamafile combines llama.cpp with Cosmopolitan Libc in a single executable.

How to use

  1. Download the llamafile.

  2. Open a Linux terminal.

  3. Grant execute permission.

chmod +x Qwen3.8-27B-heretic-ara.Q4_K_M.llamafile
  1. Run a text prompt with thinking disabled.
prompt='<|im_start|>user
What is the capital of France? Answer with only the city name.<|im_end|>
<|im_start|>assistant
<think>

</think>

'
/bin/sh Qwen3.8-27B-heretic-ara.Q4_K_M.llamafile \
  --cli --no-display-prompt --no-conversation --log-file /dev/null \
  -ngl 999 -c 512 -n 64 --temp 0 -p "$prompt"
  1. Press Control-C to stop generation.

Settings for Trohrbaugh Heretic ARA GGUF Llamafiles

  • Model creator: trohrbaugh
  • Original model: trohrbaugh/Qwen3.8-27B-heretic-ara
  • Quantized GGUF files used: mradermacher/Qwen3.8-27B-heretic-ara-GGUF
    • Commit hash f69c978f262c983cb312395d18bbce340260e89f
    • Commit message: auto-patch README.md
  • Llamafile version: 0.10.5, mozilla-ai/llamafile
    • Commit hash 486e6c5f9356eae50b851b07517bfae1f2420193
    • Commit message: docs(support): document the Vulkan GPU backend (#1029) (#1033)
  • Quantizations: Q3_K_M, Q4_K_M, Q5_K_M, Q6_K, Q8_0.
  • Validation: Linux text generation; details in validation/.
  • Vision: text GGUF; upstream projector files remain separate.
  • MTP: source tensors retained; validation uses standard decoding.
  • .args contents:
-m
Qwen3.8-27B-heretic-ara.Q4_K_M.gguf
...

(Following is original model card for Trohrbaugh Heretic ARA GGUF)


Excerpt from the original GGUF source card: mradermacher/Qwen3.8-27B-heretic-ara-GGUF.

static quants of https://huggingface.co/trohrbaugh/Qwen3.8-27B-heretic-ara

Catalog is the map. Apps are the tools.

Run models on your own machine, not in the cloud.

Every model page has an "Open in app" button that hands off directly to a local runtime of your choice - Infrahuman, LM Studio, or Ollama. No API keys, no subscription, no prompt leakage.