All credits belong to Luffy the Fox

This is just a backup copy from his work!

LuffyTheFox/Qwen3.6-35B-A3B-Uncensored-Genesis-Hermes-V5-GGUF

⚡ Why Genesis project exists? Here link that explain everything.

⚡ 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 -> Genesis Hermes V5

What is Genesis? Genesis is post training data regeneration and calibrarion algorythm for neural networks (LLM) in GGUF format that I made with AI help during almost half a year of development. It's optimized, architecture independent, works with any model and based on mathematical statistics. I don't train or finetune models, I repair purity of signal in them instead on Google Collab Free on Tesla T4 GPU via Python based on how models learns information. On first stage I scan ssm_conv1d tensors in model, they handle long context memory. I repair balance in ssm_conv1d tensors via custom SVD. On second stage I scan model and detect noise in tensors via custom SVD. During scanning I exclude ssm_conv1d, token_embd.weight, output.weight, ffn_gate_inp.weight, ffn_gate_inp_shexp.weight, 1D tensors, bias and norms. Then I reduce training noise in tensors via custom SVD with preserved training data, 99% of siginal and learned gradient. On third stage, I scan blocks in model via chunks via 3 parameters and pick best one that fits to weight distribution in tensor. Best picked chunk replaces zero chunks in broken tensor without touching learned structure in model

Model is based on HauhauCS/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive base.

And DJLougen/hermes-qwen3.5-35b-a3b-GGUF finetune for Hermes agent.

I transferred data from finetune on Hermes dataset (around 2k blocks from two FFN expert tensors) to HauhauCS uncensored base.

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_K_P

Tensor drift repair by me. Method: Genesis

Links:

• Original uncensored model
• Quantization Script with Unsloth profiles support

---

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.
• Zero blocks: zero blocks corrupt the signal, turning training into noise amplification.
• Training Noise: training noise increase randomness and ruins model output quality.

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: luffythefox@mail.ru

My Telegram: @LuffyTheFox

Recommended Settings for RTX 3060 12 GB for best perfomance on APEX quant

Chat template: chat_template.jinja

Set K Cache Quantization Type and V Cache Quantization Type to F16.

Set Number of layers for which to force MoE weights onto CPU to 40.

Set GPU offload to 15. Set number of active experts to 8.

For best model stability and first experience I recommend starting from this string in your System Prompt with enabled thinking and nothing else:

You are Qwen, a large language model created by Tongyi Lab team from Alibaba Group. You are a helpful assistant.

or this string (for roleplay, add anything you want after it)

You are a helpful assistant.

If you want to bring more creativity to model use this System Prompt from discussion: link

Thinking mode (coding):

• Coding/precise tasks: temperature=0.6, top_p=0.95, top_k=20, min_p=0, seed=42, presence_penalty=disabled, repeat_penalty=disabled
• General: temperature=1.0, top_p=0.95, top_k=20, min_p=0.05, seed=42, presence_penalty=disabled, repeat_penalty=disabled

Non Thinking mode (creative):

• General: temperature=0.7, top_p=0.8, top_k=20, min_p=0.0, seed=42, presence_penalty=disabled, repeat_penalty=disabled

For agentic tasks you can use this System Prompt:

You are Qwen, a large language model created by Tongyi Lab team from Alibaba Group. You are a helpful assistant that answers in JSON. Here's the json schema you must adhere to:\n<schema>\n{schema}\n</schema>.

With those settings:

• General: temperature=0.7, top_p=0.8, top_k=20, min_p=0.0, seed=42, presence_penalty=disabled, repeat_penalty=disabled, thinking=disabled

And commands from this dataset: hermes-function-calling-v1

Testing

2D animation testing

System Prompt: You are a helpful assistant.

Settings: temperature=1.0, top_p=0.95, top_k=20, min_p=0.05, seed=42, presence_penalty=disabled, repeat_penalty=disabled

Prompt 1: Generate an animated SVG on animated background of a Pingu waving on an iceberg wearing his iconic winter scarf.

Prompt 2: Animate his wings and fix floating wing

Result: pingu_animated.svg

Static 2D testing

System Prompt: You are Qwen, a large language model created by Tongyi Lab team from Alibaba Group. You are a helpful assistant.

Settings: temperature=0.6, top_p=0.95, top_k=20, min_p=0, presence_penalty=disabled, repeat_penalty=disabled

Prompt: Generate an SVG of a pelican riding a bicycle

Result: pelican.svg

On next stage I asked model: Replace pelican with rooster

Result: rooster.svg

I asked model: Replace rooster with cock

Result: cock.svg

Finally I asked model: Replace rooster with Pingu

Result: pingu.svg

Important:

• Keep at least 128K context to preserve thinking capabilities
• Use --jinja flag with llama.cpp for proper chat template handling
• Vision support requires the mmproj file alongside the main GGUF

---

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)
• 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.