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

Concave arm-to-reward function

Benchmarks

Task NameDataset NameSOTA ResultTrend
Best Arm IdentificationConcave arm-to-reward function (K=40) (synthetic)
Error Rate0.02
15
Best Arm IdentificationConcave arm-to-reward function K=20 (synthetic)
Error Probability0.09
15
Best Arm IdentificationConcave arm-to-reward function K=10 (synthetic)
Error Rate0.04
15
Showing 3 of 3 rows