Knowledge & Instruction Following on MMLU
88.21ScoreQwen3.5-9B (released)
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen3.5-9B (released)Model Size=9B, Training Protocol=Released, Decoding Strategy=Native AR decoding2026.06 | 88.21 | |
| Qwen3.5-9BParams=9B, Sampling mode=AR2026.06 | 88.21 | |
| LLaDA-2.0-flashParams=100B-A5B, Sampling mode=Diffusion2026.06 | 87.69 | |
| Qwen3.5-9B + AR-SFTModel Size=9B, Training Protocol=AR-SFT, Decoding Strategy=Native AR decoding2026.06 | 86.07 | |
| Qwen3.5-4B (released)Model Size=4B, Training Protocol=Released, Decoding Strategy=Native AR decoding2026.06 | 85.22 | |
| Qwen3.5Params=4B, Sampling mode=AR2026.06 | 85.22 | |
| FLARE-9BModel Size=9B, Training Protocol=FLARE, Decoding Strategy=AR-Trust sampling2026.06 | 84.8 | |
| FLARE-9BParams=9B, Sampling mode=AR-Trust2026.06 | 84.8 | |
| SDAR 30B-A3BParams=30B-A3B, Sampling mode=Diffusion2026.06 | 82.8 | |
| Qwen3.5-4B + AR-SFTModel Size=4B, Training Protocol=AR-SFT, Decoding Strategy=Native AR decoding2026.06 | 80.75 | |
| FLARE-9BParams=9B, Sampling mode=Diffusion-Trust2026.06 | 80.75 | |
| LLaDA-2.0-miniParams=16B-A1B, Sampling mode=Diffusion2026.06 | 80.53 | |
| FLAREParams=4B, Sampling mode=Diffusion-Trust2026.06 | 79.54 | |
| FLARE-4BModel Size=4B, Training Protocol=FLARE, Decoding Strategy=AR-Trust sampling2026.06 | 78.73 | |
| FLAREParams=4B, Sampling mode=AR-Trust2026.06 | 78.73 | |
| SDARParams=8B, Sampling mode=Diffusion2026.06 | 78.6 | |
| SDARParams=4B, Sampling mode=Diffusion2026.06 | 74.9 | |
| Qwen3.5-2B (released)Model Size=2B, Training Protocol=Released, Decoding Strategy=Native AR decoding2026.06 | 73.59 | |
| Qwen3.5Params=2B, Sampling mode=AR2026.06 | 73.59 | |
| Qwen3.5-2B + AR-SFTModel Size=2B, Training Protocol=AR-SFT, Decoding Strategy=Native AR decoding2026.06 | 68.63 | |
| FLARE-2BModel Size=2B, Training Protocol=FLARE, Decoding Strategy=AR-Trust sampling2026.06 | 67.6 | |
| FLAREParams=2B, Sampling mode=AR-Trust2026.06 | 67.6 | |
| FLAREParams=2B, Sampling mode=Diffusion-Trust2026.06 | 64.14 | |
| SDARParams=1.7B, Sampling mode=Diffusion2026.06 | 62.9 |