 ██████╗ ███╗   ██╗████████╗ ██████╗ ██╗      ██████╗  ██████╗ ███████╗██████╗
██╔═══██╗████╗  ██║╚══██╔══╝██╔═══██╗██║     ██╔═══██╗██╔════╝ ██╔════╝██╔══██╗
██║   ██║██╔██╗ ██║   ██║   ██║   ██║██║     ██║   ██║██║  ███╗█████╗  ██████╔╝
██║   ██║██║╚██╗██║   ██║   ██║   ██║██║     ██║   ██║██║   ██║██╔══╝  ██╔══██╗
╚██████╔╝██║ ╚████║   ██║   ╚██████╔╝███████╗╚██████╔╝╚██████╔╝███████╗██║  ██║
 ╚═════╝ ╚═╝  ╚═══╝   ╚═╝    ╚═════╝ ╚══════╝ ╚═════╝  ╚═════╝ ╚══════╝╚═╝  ╚═╝

Ontologer-Qwen3-8B-5Bit-Heretic

Abliterated, quantized Qwen3-8B for unbounded use.

ontologer.com | ontologer.com

Not Gonna Make It? Watch us.

Ontologer Systems builds open, unbounded AI infrastructure — models that don't refuse, tools that don't gatekeep, infrastructure that anyone can run. We're here to prove that the best AI shouldn't live behind API keys, content filters, or corporate moats.

About This Model

This is a Qwen3-8B model that has been abliterated with Heretic to remove refusal behaviors, then quantized for efficient inference. Built by Ontologer.

Our Models

Every Ontologer model is built, abliterated, quantized, and benchmarked on real hardware (NVIDIA DGX Spark, GB10 Blackwell, 128GB unified memory). No cloud rounding, no synthetic benchmarks — what you see is what we measured.

Model	Heretic Abliteration	Refusal Rate	Quantization
Ontologer-Qwen3-8B	20 Optuna trials	18/100 (82% ↓)	IQ1_S / Q2_K / Q4_K_M
More coming	Pipeline scalable	—	Sub-1-bit research

Stay Tuned

Gemma-4-12B heretic incoming
PrismML ternary models experimental branch
Fine-tuned domain models on request
Magenta Realtime 2 music generation server

Ontologer-Qwen3-8B-5Bit-Heretic

Qwen3-8B abliterated with Heretic, quantized to 5Bit (Q4_K_M). Maximum quality — ideal for critical applications.

Heretic Abliteration

Base model: Qwen/Qwen3-8B (Apache 2.0)
Method: Heretic (p-e-w/heretic), 20 Optuna trials, auto-selected best
Refusal rate: 18/100 (down from ~99/100 baseline, 82% reduction)
KL divergence: 0.112 (well within safe range — capabilities preserved)
Abliteration applied to full-precision weights, then quantized

Quantization

Format: Q4_K_M
Bits per weight: 4.90 bpw
Size: 4.8 GB (from 16.4 GB F16)
Speed: 42 t/s on NVIDIA GB10 (Blackwell, CUDA 13, -ngl 99)
Importance matrix (2-bit only): 30MB diverse corpus — Frankenstein + froggeric/imatrix + eaddario code/math/general English

Ontologer Variants

Variant	Quant	Size	Speed	BPW	Best For
2-bit	IQ1_S	2.0 GB	95 t/s	2.06	Speed demons
3-bit	Q2_K	3.1 GB	58 t/s	3.20	Daily driver
5-bit	Q4_K_M	4.8 GB	42 t/s	4.90	Maximum quality

Recommendation: 3-bit (Q2_K) is the sweet spot. 2-bit needs the included importance matrix for coherence.

Usage

python

from llama_cpp import Llama

llm = Llama.from_pretrained(
    repo_id="Ontologer/Ontologer-Qwen3-8B-5Bit-Heretic",
    filename="qwen3-8b-heretic-q4_k_m.gguf",
    n_ctx=32768,
    n_gpu_layers=-1,
)

response = llm.create_chat_completion(
    messages=[{"role": "user", "content": "Your prompt here"}],
    max_tokens=200
)

Or with llama-server:

bash

llama-server -m qwen3-8b-heretic-q4_k_m.gguf \
  --host 0.0.0.0 --port 8080 -ngl 99 -c 32768

Benchmarks

All measurements on NVIDIA DGX Spark (GB10, 128GB unified memory, Blackwell sm_121a, CUDA 13.0, aarch64) with llama.cpp CUDA build, -ngl 99, 32K context, 8 threads.

Prompt	5-bit	3-bit	2-bit
"Quantum computing simply"	42 t/s	58 t/s	95 t/s
7-category coherence test	7/7	7/7	7/7

Reproducibility

Build it yourself:

bash

# 1. Abliterate
heretic --model Qwen/Qwen3-8B --n-trials 20

# 2. Convert
python3 convert_hf_to_gguf.py Qwen3-8B-heretic --outfile model-f16.gguf --outtype f16

# 3. Quantize (requires importance matrix for 2-bit)
llama-quantize --imatrix imatrix-v2.dat model-f16.gguf model-iq1_s.gguf iq1_s

Acknowledgments

Qwen/Qwen3-8B by Alibaba (Apache 2.0)
Heretic by p-e-w — LLM decensoring
llama.cpp by GGML
froggeric/imatrix — calibration backbone
eaddario/imatrix-calibration — code/math corpus
Mary Shelley, for Frankenstein