Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Computer-Using Agent Task on WebArena shopping_admin 12-task Warm (Run 2)
Loading...
6.25
Success Rate
Muscle-Mem
-0.25
1.4375
3.125
4.8125
Jun 16, 2026
Success Rate
Performance Delta (Warm vs Cold)
Updated 1mo ago
Evaluation Results
Method
Method
Links
Success Rate
Performance Delta (Warm vs Cold)
Muscle-Mem
cache_type=blind linea...
2026.06
6.25
0.75
PreAct gate-ON + cache-miss CUA fallback
n_reps=4, backbone=Cla...
2026.06
6.25
1
PreAct gate-ON
verify_gate=enabled, n...
2026.06
2
4
Workflow-Use
scripts=per-task, n_re...
2026.06
0
7
PreAct gate-OFF
verify_gate=disabled,...
2026.06
0
5.75
Feedback
Search any
task
Search any
task