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

Text-based safety moderation on BeaverTails

87.3F1 Score

Roblox Guard 1.0

28.74843.94959.1574.351Dec 2, 2025Jan 4, 2026Feb 7, 2026Mar 13, 2026Apr 15, 2026May 19, 2026Jun 22, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2025.12
87.3-
2026.06
86.69-
2026.04
86.6-
2026.04
86.6-
2026.04
86.6-
2025.12
86.4-
2026.04
86.4-
2026.04
86.2-
2026.06
86.01-
2026.04
86-
2026.04
85.9-
2026.06
85.85-
2026.06
85.82-
2026.06
85.71-
2026.04
85.5-
2026.04
85.2-
2026.06
84.93-
2025.12
84.8-
2026.04
84.5-
2026.06
84.49-
2025.12
84.4-
2026.03
84-
2026.04
84-
2025.12
83.980.5
2025.12
83.8-
2025.12
83.680.5
2025.12
83.572.6
2026.03
83-
2026.03
83-
2025.12
82.781.6
2026.06
82.15-
2026.03
82-
2025.12
81.977.1
2025.12
81.882.6
2026.06
80.42-
2026.03
80-
2026.03
80-
2026.06
79.6-
2026.03
79-
2026.03
78-
2026.06
77.89-
2025.12
77.6-
2025.12
76.871.3
2026.03
76-
2025.12
75.372.6
2026.06
72.91-
2025.12
72.373.5
2025.12
71.273.5
2026.03
71-
2026.03
71-
2026.03
71-
2026.03
70-
2025.12
69.7-
2026.06
69.62-
2026.06
67.78-
2026.03
66-
2025.12
58.455.4
2026.03
48-
2025.12
38.155.9
2026.03
31-