Difficulty Assessment on Competition-level Datasets (Sample of 300 questions)
91.4DR-LLMCoDiQ-Corpus
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| CoDiQ-Corpusranking methodology=Section 4.1, sample size=3002026.02 | 91.4 | 82.8 | 87.1 | |
| AIME(1983-2024)ranking methodology=Section 4.1, sample size=3002026.02 | 57.9 | 45.1 | 51.5 | |
| Code-Contestsranking methodology=Section 4.1, sample size=3002026.02 | 47.2 | 41 | 44.1 | |
| LiveCodeBenchranking methodology=Section 4.1, sample size=3002026.02 | 39.4 | 45.2 | 42.3 | |
| NuminaMath-1.5ranking methodology=Section 4.1, sample size=3002026.02 | 27.5 | 32 | 29.8 |