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

Agent Task on SciWorld (Suc@1, Suc@2, Suc@3)

98.4Success Rate @1

PivoARL

5.73629.79353.8577.907Jul 4, 2026
Updated 18d ago

Evaluation Results

MethodLinks
2026.07
98.4100100
2026.07
94.597.698.4
2026.07
89.894.596
2026.07
83.594.598.4
2026.07
82.18989.8
2026.07
81.790.292.6
2026.07
79.691.493.2
2026.07
78.986.789.8
2026.07
75.882.886.7
2026.07
42.955.460.9
2026.07
32.842.147.6
2026.07
9.412.512.5
2026.07
9.314.717.9