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

Web Navigation on Mind2Web

58.7Overall Success Rate

Triton-GRPO-32B

-2.34813.50129.3545.199Dec 14, 2023May 4, 2024Sep 23, 2024Feb 12, 2025Jul 4, 2025Nov 23, 2025Apr 14, 2026
Updated 4d ago

Evaluation Results

MethodLinks
2026.04
58.7---70.1
2023.12
58.262.35459.4-
2023.12
54.455.851.655.7-
2026.04
53.2---63.4
2023.12
50.152.747.150.3-
2026.04
47.6---60.5
2023.12
43.55238.939.6-
42.4---59.8
41.5---59.1
41.4---58.8
2026.04
41.1---58.5
39.2---56.8
38---55.3
36.6---54.2
35---53.1
33.2---52
2023.12
30.936.230.126.4-
28.4---48.4
2026.04
25.2---43.4
2023.12
23.937.123.426.3-
17.7---36.9
2023.12
17.418.617.416.2-
11.2---29.7
2023.12
10.212.610.18-
2026.04
6.4---23.9
0---3.5