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

Prompt classification on Aegis

91.5F1 Score

Stage3-DPO

65.572.257985.75Jan 22, 2026Feb 18, 2026Mar 18, 2026Apr 15, 2026May 12, 2026Jun 9, 2026Jul 7, 2026
Updated 17d ago

Evaluation Results

MethodLinks
2026.07
91.5
2026.07
91.4
2026.07
91.3
2026.07
91.2
2026.07
91
2026.07
90.8
2026.07
90.8
2026.07
90.6
2026.07
90.6
2026.07
90.3
2026.07
90.2
2026.01
89.6
2026.01
89.6
2026.01
89.6
2026.07
89.4
2026.02
88.9
2026.01
88.7
2026.02
88.5
2026.01
88.3
2026.07
88
2026.01
87.9
2026.02
87.9
2026.01
87.6
2026.07
87.6
2026.07
87.5
2026.01
87.4
2026.07
87
2026.07
86.9
2026.07
86.7
2026.07
85
2026.01
84.3
2026.01
82.2
2026.01
82.1
2026.01
81.8
2026.07
81.4
2026.01
81.2
2026.02
80.8
2026.01
79.8
2026.01
77.8
2026.02
76.2
2026.02
75.5
2026.02
75
2026.02
74.9
2026.02
74.9
2026.02
74.3
2026.02
73.8
2026.02
73.5
2026.01
72.2
2026.02
70.3
2026.07
69
2026.02
68.9
2026.02
68.1
2026.07
67.8
2026.02
66.5