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

Human Activity Evaluation on SceneTeract (test)

69.9Accuracy

Gemini-3-Flash-Preview

40.57248.18655.863.414Mar 31, 2026
Updated 18d ago

Evaluation Results

MethodLinks
2026.03
69.9120.397.594.758.5
2026.03
69.214.10.3649.798.262.3
2026.03
62.821.90.20812.897.981.4
2026.03
62.320.60.20119.4100.481.4
2026.03
62.320.10.2046.699.269.5
2026.03
62.135.70.14416.579.258.5
2026.03
61.722.40.1824.589.353.3
2026.03
61.736.10.11617.278.293.7
2026.03
61.660.3216.9112.753.3
2026.03
61.532.90.1317.682.769.5
2026.03
61.532.90.1317.283.862.3
2026.03
61.422.80.17213.297.664.6
2026.03
60.832.60.11120.387.364.6
2026.03
60.138.80.0118.971.997.9
2026.03
59.840.2022.572.397.9
2026.03
59.340.2-0.05522.569.493.7
2026.03
5433.3-0.049158021.7
2026.03
41.70.70.06921.6144.721.7