#ModelAccuracyPaper
1Removing Non-English Answers48.60link
2GPT Verification48.30link
3Removing Long Paragraphs46.80link
4Random Filtering46.70link
5Shortest Answers Selection46.70link
6No Filtering (not compute-controlled)46.40link
7EASY-EP23.60link
8TENP20.20link
Science science Leaderboard