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

Wide Baseline Matching on ReasonMatch-Bench Overall (test)

70.5F1 Score

Qwen3-VL-8B + DCRL

25.7837.394960.61Jun 2, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
70.570.371.1
2026.06
57.956.959.4
2026.06
51.550.652.8
2026.06
49.250.748.7
42.842.443.4
41.743.741.1
34.834.235.5
33.737.232.2
33.532.734.7
2026.06
27.527.129.1