YOYO-AI/Qwen3.6-35B-A3B-YOYO-V2

Merge the official Qwen 35B-A3B models using the most advanced data-free merging algorithms available today.

Model Highlights:

• —*Merge Method: SWUDI-A*
• —*Precision: dtype: bfloat16*
• —*Context Length: 262,144*

Parameter Settings:

Non-Thinking Mode: (`{%- set enable_thinking = false %}`)

General Tasks:

[!TIP] `temperature=0.7`, `top_p=0.8`, `top_k=20`, `min_p=0.0`, `presence_penalty=1.5`, `repetition_penalty=1.0`

Reasoning Tasks:

[!NOTE] `temperature=1.0`, `top_p=1.0`, `top_k=40`, `min_p=0.0`, `presence_penalty=2.0`, `repetition_penalty=1.0`

Thinking Mode: (`{%- set enable_thinking = true %}`)

General Tasks:

[!TIP] `temperature=1.0`, `top_p=0.95`, `top_k=20`, `min_p=0.0`, `presence_penalty=1.5`, `repetition_penalty=1.0`

Coding Tasks:

[!NOTE] `temperature=0.6`, `top_p=0.95`, `top_k=20`, `min_p=0.0`, `presence_penalty=0.0`, `repetition_penalty=1.0`

Papers:

[https://arxiv.org/abs/2606.07289v1](https://arxiv.org/abs/2606.07289v1)

Some Details:

The specific merging details are as follows:

1. Initialization starting point: Simple averaging

2. Layer-wise rank truncation rule: Participation-Square-Root Rule