#ModelAccuracyPaper
1MetaAPO61.43link
2SELM61.01link
3Online DPO60.24link
4DPO59.39link
5SFT57.59link
6Base OPM + XKD z (ours)55.12link
7Base OPM + XKD h (ours)54.42link
8Chat – (original)54.35link
9Base – (original)54.27link
10Chat OPM + XKD z (ours)54.27link
11Chat OPM + XKD h (ours)54.01link
12Base OPM + UT52.61link
13Chat OPM + UT52.22link
14Panda-3B35.20link
15Surefire-3B33.90link
16LLaMA-3.2-3B33.30link
17Surefire-1B30.20link
18LLaMA-3.2-1B29.80link
19Panda-1B28.90link
20Base OPM26.02link
21Chat OPM25.17link
ARC arc Leaderboard