Scientific Question Answering on GPQA-D (Accuracy)
14.43Accuracy (GPQA-D)IOA
Evaluation Results
| Method | Links | |
|---|---|---|
| IOAStudent Model=Qwen2.5-3B, Teacher Model=DeepSeek-R12026.02 | 14.43 | |
| IOAStudent Model=LLaMA3.2-3B, Teacher Model=DeepSeek-R12026.02 | 12.47 | |
| MADAStudent Model=Qwen2.5-3B, Teacher Model=DeepSeek-R12026.02 | 12.27 | |
| Star-AgentsStudent Model=Qwen2.5-3B, Teacher Model=DeepSeek-R12026.02 | 11.92 | |
| CasCoDStudent Model=Qwen2.5-3B, Teacher Model=DeepSeek-R12026.02 | 11.54 | |
| CounterDistillStudent Model=Qwen2.5-3B, Teacher Model=DeepSeek-R12026.02 | 11.29 | |
| DSSStudent Model=Qwen2.5-3B, Teacher Model=DeepSeek-R12026.02 | 11.18 | |
| MADAStudent Model=LLaMA3.2-3B, Teacher Model=DeepSeek-R12026.02 | 11.13 | |
| Star-AgentsStudent Model=LLaMA3.2-3B, Teacher Model=DeepSeek-R12026.02 | 10.94 | |
| CasCoDStudent Model=LLaMA3.2-3B, Teacher Model=DeepSeek-R12026.02 | 10.72 | |
| LionStudent Model=Qwen2.5-3B, Teacher Model=DeepSeek-R12026.02 | 10.62 | |
| CounterDistillStudent Model=LLaMA3.2-3B, Teacher Model=DeepSeek-R12026.02 | 10.51 | |
| DSSStudent Model=LLaMA3.2-3B, Teacher Model=DeepSeek-R12026.02 | 10.37 | |
| LionStudent Model=LLaMA3.2-3B, Teacher Model=DeepSeek-R12026.02 | 9.93 | |
| LaMiniStudent Model=Qwen2.5-3B, Teacher Model=DeepSeek-R12026.02 | 9.63 | |
| Self-InstructStudent Model=Qwen2.5-3B, Teacher Model=DeepSeek-R12026.02 | 9.26 | |
| LaMiniStudent Model=LLaMA3.2-3B, Teacher Model=DeepSeek-R12026.02 | 8.94 | |
| Self-InstructStudent Model=LLaMA3.2-3B, Teacher Model=DeepSeek-R12026.02 | 8.52 | |
| UndistilledStudent Model=Qwen2.5-3B, Teacher Model=None2026.02 | 7.94 | |
| UndistilledStudent Model=LLaMA3.2-3B, Teacher Model=None2026.02 | 7.12 |