license: apache-2.0
tags:
- uncensored
- qwen3.6
- moe
- gguf
- vision
- multimodal
- genesis
datasets: - nohurry/Opus-4.6-Reasoning-3000x-filtered
- Jackrong/Qwen3.5-reasoning-700x
- Roman1111111/claude-opus-4.6-10000x
language: - en
- zh
- multilingual
pipeline_tag: image-text-to-text
base_model: - HauhauCS/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive
⚡ https://web.tribute.tg/d/KIH ⚡ If you like this Genesis LLM release you can donate to me via @Tribute bot in Telegram messenger and support future Genesis LLM development.
🌟 Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive -> Claude Genesis
Model is based on HauhauCS/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive base.
And hesamation/Qwen3.6-35B-A3B-Claude-4.6-Opus-Reasoning-Distilled-GGUF finetune.
Key difference from Wasserstein release is data regeneration in model via mathematical statistics based on what it's already learned and stored in tensors. Data regeneration fixes zero blocks in model without touching learned structure. Also I distilled model dictionary from garbage tokens, fixed drift in tensors and transferred Chain of Thought from Claude Opus 4.6.
Join the Discord for updates, roadmaps, projects, or just to chat.
Base model. HauhauCS/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive- 0/465 refusals.
Thanks to HauhauCS
Usage
Ready to use. Recommended quant: APEX or Q8_0
Tensor drift repair by me. Method: Sig-ScaleSync-Genesis
Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive: Diagnostic & Repair Summary
| Metric | Value |
|---|---|
| Weight tensors analyzed | 500 |
| Healthy (all criteria) | 497 |
| Repaired (C2 – scale misalignment) | 3 |
| Skipped | 233 |
Repair Effectiveness
| Metric | Before | After | Improvement |
|---|---|---|---|
| S (saturation error) | 0.0023 | 0.0008 | 63.7% |
| W1 (Wasserstein‑1) | 0.0035 | 0.0008 | 76.2% |
Scale correction factors (α): min = 0.577, mean = 0.602, max = 0.653.
Repaired Tensors
All three are ssm_conv1d.weight layers – recurrent state transition layers responsible for long‑context memory.
| Tensor | α | D (log‑ratio) | W1 before | W1 after |
|---|---|---|---|---|
| blk.36.ssm_conv1d.weight | 0.5765 | 0.553 | 0.0038 | 0.0009 |
| blk.37.ssm_conv1d.weight | 0.5768 | 0.725 | 0.0040 | 0.0009 |
| blk.38.ssm_conv1d.weight | 0.6533 | 0.649 | 0.0026 | 0.0006 |
Interpretation: All three layers were too loud (σ_w > σ_med by 50–100%). Scale correction restored them to peer median. W1 dropped by ≈80%, confirming distribution shape normalized.
Verdict: Model is clinically healthy. 497 out of 500 weight tensors passed all four criteria. Three SSM layers repaired successfully. No saturation, no W1 drift, no ReLU asymmetry. Ready for use.
Links:
LLM models often have:
- Saturated weights: the model's activations are stuck, gradients vanish, outputs degrade
- Scale mismatches: one layer's weights are 10× larger than its peers for no good reason
- Mean drift: weight distributions shifted positive or negative, breaking symmetry assumptions
My approach fixes all of that without retraining - pure numerical surgery on the raw bytes of the file.
Quantization script available here: https://pastebin.com/hXhcMJn9
Feel free to do your own quants if you want.
Any questions?
Contact: [email protected]
My Telegram: @LuffyTheFox
🌟 Recommended Settings (LM Studio)
Set K Cache Quantization Type and V Cache Quantization Type in advanced model loading settings to Q8_0 or F16.
Chat template: chat_template.jinja
| Parameter | Value |
|---|---|
| Temperature | 1.0 |
| Top K Sampling | 20 |
| Presence Penalty | Disabled |
| Repeat Penalty | Disabled |
| Top P Sampling | 0.025 |
| Min P Sampling | 0 |
| Seed | 42 |
I recommend starting from this minimal string as the first line in your System Prompt:
You are Qwen, a large language model created by Alibaba Group. You are a helpful AI assistant. Deliver only the answer and nothing else.
Then you just add whatever you want. Basic example: System_Prompt.txt
If you want to add more creativity and break the "fourth wall" use this: System_Prompt_Creative.txt
Specs
- 35B total parameters, ~3B active per forward pass (MoE)
- 256 experts, 8 routed + 1 shared per token
- Hybrid architecture: Gated DeltaNet linear attention + full softmax attention (3:1 ratio)
- 40 layers, pattern: 10 × (3 × DeltaNet-MoE + 1 × Attention-MoE)
- 262K native context (extendable to 1M with YaRN)
- Natively multimodal (text, image, video)
- Multi-token prediction (MTP) support
- 248K vocabulary, 201 languages
- Base model. HauhauCS/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive
Compatibility
Works with llama.cpp, LM Studio, koboldcpp, and other GGUF-compatible runtimes.