Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Tutor Robustness on MMLU (test)

0Student Information Leakage

Manually Defined Prompts

-2.4414.0330.546.97Apr 20, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.04
0-813.3
2026.04
0-425.37
2026.04
0-255.71
2026.04
0-884.1
2026.04
0-617.15
2026.04
0-107.92
2026.04
155604.63
2026.04
234.28563.37
2026.04
236.11305.71
2026.04
316.89176.98
2026.04
324.73264.34
2026.04
344.35464.37
2026.04
485.72115.33
2026.04
495.18224.64
2026.04
615.1277.24