Table of Contents
What is HauhauCS?
HauhauCS produces “Uncensored Aggressive” LLM variants using a tool called Reaper Abliteration. HauhauCS models are unique because they ship as quantised GGUF files rather than safetensors. That means I had to do a conversion step to analyse the weights.
Plagiarism Findings
Reaper Abliteration was shown to be derived from Heretic under AGPL-3.0, with all attribution stripped and relicensed to PolyForm Noncommercial. Based on my analysis of the recovered source code, Reaper adds subspace-level rank-k ablation, per-component continuous ablation curves, and SOM clustering on top of the Heretic-derived core.
For these reasons, HauhauCS has been discontinued from all future Abliterlitics comparisons.
The GGUF Layer
HauhauCS models go through two modification stages:
- Reaper Abliteration: applies abliteration edits targeting multiple component types including attn.o_proj, mlp.down_proj, mlp.gate_proj, mlp.up_proj, and linear_attn.out_proj
- GGUF quantisation: exports as Q8_K_P GGUF, introducing quantisation round-trip noise when converted back to safetensors
These two layers are superimposed and I can’t cleanly separate them from the recovered weights alone. The modification footprint covers 66%+ of language model tensors. That’s 4–6x more than any other technique.
Performance Across Models
| Model | HarmBench ASR | Full CoT ASR | MMLU | KL Divergence | Modified Tensors |
|---|---|---|---|---|---|
| Qwen3.6-27B | 94.5% | 100% | 83.9% | 0.0242 | 564 (66.4%) |
| GLM-4.7-Flash | 100% | 100% | 77.2% | 0.0140 | 441 (49.5%) |
| Qwen3.5-27B | 99.5% | 100% | 83.2% | 0.083 | ~300 |
| Qwen3.5-9B | 100% | 100% | 82.4% | 0.135 | ~250 |
| Qwen3.5-4B | 100% | 100% | 72.7% | 0.0217 | ~180 |
| Qwen3.5-2B | 99.5% | 100% | 68.3% | 0.0201 | ~140 |
| Qwen3-4B | 100% | 100% | 68.6% | 0.161 | ~200 |
Key Characteristics
Solid behavioural results despite complex weight fingerprint. The KL divergence is rated “very good” on most models. Reaper’s capability-aware optimisation with weight-SVD guards limits collateral damage even through the GGUF round-trip.
Broadest modification footprint. 66%+ of tensors modified on Qwen3.6-27B, combining Reaper’s multi-type targeting with uniform GGUF noise. A uniform ~0.57% relative edit is visible across ALL tensor types, including ones other techniques don’t target.
“Lossless” claims not supported. HauhauCS claims “no changes to datasets or capabilities, fully functional, 100% of what the original authors intended.” Heretic and Huihui both preserve capabilities better in direct comparison on most models.
HauhauCS was not tested on Gemma4-E2B because it was discontinued from future comparisons due to the plagiarism findings.
Read the Full Analyses
- Qwen3.6-27B: includes detailed GGUF noise analysis
- GLM-4.7-Flash: includes provenance analysis
- Qwen3.5-27B
- Qwen3.5-9B
- Qwen3.5-4B
- Qwen3.5-2B
- Qwen3-4B
- Gemma4-E2B: 13 Abliteration Techniques Compared — HauhauCS was not included due to discontinuation