Question Answering on Riddle (test) benchmark leaderboard · SOTA2 Research

Question Answering on Riddle (test)

67.84AccuracyOracle

Evaluation Results

Method	Links
Oracle2026.04		67.84
REALM2026.04		65.45
REALM2026.04		62.31
NOISY2026.04		61.49
REALM2026.04		60.43
NOISY2026.04		57.06
Oracle2026.04		54.12
NOISY2026.04		50.35
REALM2026.04		49.61
REALM2026.04		46.63
REALM2026.04		46.04
NOISY2026.04		45.84
NOISY2026.04		42
Oracle2026.04		39.02
Oracle2026.04		39.02
NOISY2026.04		36.2
REALM2026.04		34.71
REALM2026.04		34.71
REALM2026.04		32.16
REALM2026.04		31.49
noisy2026.04		31.25
NOISY2026.04		31.25
NOISY2026.04		29.69
NOISY2026.04		23.8
Oracle2026.04		0.6784
REALM2026.04		0.6231
REALM2026.04		0.6133
REALM2026.04		0.6125
NOISY2026.04		0.5792
NOISY2026.04		0.5706
NOISY2026.04		0.5616
Oracle2026.04		0.5412
REALM2026.04		0.4694
REALM2026.04		0.4663
REALM2026.04		0.4647
NOISY2026.04		0.4247
NOISY2026.04		0.4208
NOISY2026.04		0.42
Oracle2026.04		0.3902
REALM2026.04		0.3259
REALM2026.04		0.3227
REALM2026.04		0.3216
NOISY2026.04		0.2969
NOISY2026.04		0.291
NOISY2026.04		0.2816