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

Preference Modeling on RewardBench 2

68.42Factuality

HRC

42.149648.969855.7962.6102May 17, 2026
Updated 15d ago

Evaluation Results

MethodLinks
2026.05
68.423560.1192.8994.7574.5170.95
2026.05
6832.561.292.6795.1574.5170.67
2026.05
67.5833.1257.9292.2293.7473.5369.69
2026.05
66.5333.7562.392.2294.1471.5770.08
2026.05
64.6334.3864.4892.6790.9173.5370.1
2026.05
47.5835.6361.758279.639.2257.63
2026.05
47.1633.7562.847871.9238.2455.32
2026.05
45.8933.7562.383.7877.7839.2257.12
2026.05
45.6832.562.380.6777.1737.2555.93
2026.05
43.1636.2564.4881.1176.1637.2556.4