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

Active Perception on HR-Bench 8K

85.5Accuracy

Gemini-3-Flash

61.78867.94474.180.256Jun 30, 2026
Updated 23d ago

Evaluation Results

MethodLinks
2026.06
85.5
2026.06
83.15
2026.06
81.62
2026.06
79.88
2026.06
74.63
2026.06
72.88
2026.06
72
2026.06
69.5
2026.06
67.5
2026.06
66.9
2026.06
62.7