Difficulty Correlation with Human Performance on CMCQRD n=787
0.53Pearson CorrelationLLM compare
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| LLM compareModel=Gemini 2.5 Pro2025.12 | 0.53 | 0.56 | 0.38 | |
| LLM compareModel=OpenAI o32025.12 | 0.48 | 0.51 | 0.35 |
| Method | Links | |||
|---|---|---|---|---|
| LLM compareModel=Gemini 2.5 Pro2025.12 | 0.53 | 0.56 | 0.38 | |
| LLM compareModel=OpenAI o32025.12 | 0.48 | 0.51 | 0.35 |