Physics Reasoning on OlyBench Phy
4.5Acceptance LengthTTS
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| TTSTarget Model=Qwen/Qwen3-4B2026.05 | 4.5 | 48.1 | |
| TTSTarget Model=Qwen/Qwen3-8B2026.05 | 4.3 | 53.1 | |
| TTSTarget Model=Qwen/Qwen3.5-122B2026.05 | 3.9 | 28.4 | |
| DFlashTarget Model=Qwen/Qwen3-4B2026.05 | 3.1 | — | |
| DFlashTarget Model=Qwen/Qwen3.5-122B2026.05 | 3 | — | |
| DFlashTarget Model=Qwen/Qwen3-8B2026.05 | 2.8 | — | |
| TTSModel=Llama3.1-8B2026.05 | 1.9 | 65.4 | |
| TTSModel=Qwen/Qwen3-8B2026.05 | 1.9 | 17.2 | |
| TTSModel=Qwen/Qwen3-32B2026.05 | 1.8 | 21.6 | |
| EAGLE-3Model=Qwen/Qwen3-8B2026.05 | 1.6 | — | |
| EAGLE-3Model=Qwen/Qwen3-32B2026.05 | 1.5 | — | |
| EAGLE-3Model=Llama3.1-8B2026.05 | 1.2 | — |