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

Personalized Reward Modeling on PRISM Personalized

68.06Accuracy

P-GenRM

50.17254.81659.4664.104Jan 6, 2026Jan 12, 2026Jan 18, 2026Jan 24, 2026Jan 30, 2026Feb 5, 2026Feb 12, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.02
68.06
2026.02
67.54
2026.02
66.21
2026.02
65.32
2026.01
65.11
2026.02
63.74
2026.02
63.44
2026.02
63.27
2026.02
62.84
2026.01
62.74
2026.02
61.61
2026.01
61.39
2026.01
60.91
2026.02
59.7
2026.02
59.16
2026.02
58.33
2026.01
58.26
2026.02
58.25
2026.01
58.23
2026.02
58.19
2026.02
57.75
2026.02
57.29
2026.02
57.11
2026.01
56.81
2026.02
56.74
2026.01
56.48
2026.02
56.24
2026.01
55.47
2026.01
55.26
2026.01
55.24
2026.02
54.7
2026.01
54.23
2026.02
54.21
2026.01
54.17
2026.02
54.09
2026.02
54.02
2026.01
53.36
2026.02
52.96
2026.01
52.8
2026.02
52.58
2026.01
52.09
2026.02
52.04
2026.01
51.65
2026.01
50.86