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

Active Perception on HR-Bench-4K

89.25Accuracy

Gemini-3-Flash

67.35873.041578.72584.4085Jun 30, 2026
Updated 23d ago

Evaluation Results

MethodLinks
2026.06
89.25
2026.06
85
2026.06
84.5
2026.06
81.75
2026.06
78.88
2026.06
78.25
2026.06
77
2026.06
74
2026.06
73.2
2026.06
71.75
2026.06
68.2