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

Vision-Language-Action instruction following on ICBench Goal Suite (test)

96.2Success Rate (SR)

π0.5

34.00850.15466.382.446Mar 6, 2026
Updated 4mo ago

Evaluation Results

MethodLinks
2026.03
96.21.4
2026.03
92.45.2
2026.03
90.27.4
2026.03
82.814.8
2026.03
70.227.8
2026.03
66.431.6
2026.03
64.233.8
2026.03
58.439.6
2026.03
46.449.4
2026.03
46.249.6
2026.03
40.255.6
2026.03
36.459.4