Reasoning and Knowledge Assessment on ARC bo
53.67AccuracyOurs-MoE-SFT
Evaluation Results
| Method | Links | |
|---|---|---|
| Ours-MoE-SFTModel=Ours-MoE-SFT2025.07 | 53.67 | |
| Ours-SFTModel=Ours-SFT2025.07 | 48.39 | |
| Ours-MoE-Base-8kModel=Ours-MoE-Base-8k2025.07 | 45.79 | |
| Ours-MoE-BaseModel=Ours-MoE-Base2025.07 | 44.96 | |
| Ours-Base-32kModel=Ours-Base-32k2025.07 | 44.62 | |
| Ours-BaseModel=Ours-Base2025.07 | 44.32 | |
| DeepSeek-R1-Distill-Llama-8BModel=DeepSeek-R1-Distill-Llama-8B2025.07 | 30.38 | |
| Qwen2.5-7B-baseModel=Qwen2.5-7B-base2025.07 | 29.6 | |
| LLaMA3.1-8B-InstructModel=LLaMA3.1-8B-Instruct2025.07 | 29.31 | |
| Qwen3-8BModel=Qwen3-8B2025.07 | 29.16 | |
| Qwen2.5-7B-InstructModel=Qwen2.5-7B-Instruct2025.07 | 28.47 |