Online A/B Test
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Online A/B test Production Traffic Four-week online deployment
16.67Revenue
5
Online A/B Test October 5, 2025 to October 12, 2025
-2.45Marketing Cost Change
1
4-week Online A/B Test
40.97Video Progress Completion Rate (25%)
1
Online A/B Test Ranking
62Engagement Rate 1
1
Online A/B Test Pre-Ranking
75Engagement Metric 1
1
Online A/B Test (Overall)
31AVDU
1
Online A/B Test Real-world traffic
2.56Few/No-Result Rate
1
Online A/B test Feed Scenario
1.29IPV
1
Online A/B Test Trinity Pipeline one-week (treatment group)
0.008AAD
1