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

Visual Question Answering on GAR-Bench-VQA

64.2Overall VQA Score

Gemini-2.5-Pro

-0.07216.61433.349.986Oct 21, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.10
64.2-62.368.858.666.7-64.164.970.3
2025.10
61.3-5870.355.263.9-54.749.271.3
2025.10
59.9-59.454.775.952.8-48.460.768.3
2025.10
53.5-34.865.348.352.8-57.860.261.4
52.8-46.45065.533.3-68.844.357.4
50.9-46.453.141.430.6-71.936.158.4
2025.10
50.6-55.146.96947.2-21.962.356.4
50.5-44.954.758.661.1-53.147.545.5
46.5-39.140.651.755.6-60.936.147.5
41.7-39.140.644.827.8-59.436.140.6
38.9-36.237.558.641.7-51.627.933.6
2025.10
38.2-55.139.141.436.1-31.336.131.7
37.5-33.32544.838.9-60.934.332.7
35.1-30.421.948.338.9-48.426.238.6
34.4-292534.530.6-43.826.244.6
2025.10
34.3-39.145.329.630.6-54.721.321.8
2025.10
2.4-2.93.16.95.6-1.61.60