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

Tabular Data Analysis on DataBench

35.6Accuracy

Weak-to-Strong (AEA)

8.24815.34922.4529.551May 22, 2026
Updated 8d ago

Evaluation Results

MethodLinks
2026.05
35.6
2026.05
35
2026.05
32.5
2026.05
31.2
2026.05
30.7
2026.05
29.7
2026.05
29.7
2026.05
29.3
2026.05
28.8
2026.05
27.7
2026.05
27.3
2026.05
27.3
2026.05
27.3
2026.05
27.3
2026.05
26.2
2026.05
23.9
2026.05
15.2
2026.05
11.7
2026.05
10.5
2026.05
9.3