Difficulty Correlation with Human Labels on Omni-Math n=1876
0.82Pearson CorrelationLLM compare
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| LLM compareModel=Gemini 2.5 Pro2025.12 | 0.82 | 0.8 | 0.62 | |
| LLM compareModel=OpenAI o32025.12 | 0.8 | 0.78 | 0.6 |
| Method | Links | |||
|---|---|---|---|---|
| LLM compareModel=Gemini 2.5 Pro2025.12 | 0.82 | 0.8 | 0.62 | |
| LLM compareModel=OpenAI o32025.12 | 0.8 | 0.78 | 0.6 |