List your company

Agreement with process human labels on CUAVerifierBench Browserbase OM2W

78AccuracyUniversal Verifier (UV)
54.0860.2966.572.71Apr 5, 2026

Evaluation results

MethodLinks
Universal Verifier (UV)Verifier=UV, Backbone=GPT-5.22026.0478570.432920
Universal VerifierVerifier=UV, Base Model=GPT-5.22026.0478570.432920
WebVoy.Verifier=WebVoy., Backbone=GPT-5.22026.0475560.42326
WebJudgeVerifier=WebJudge, Backbone=GPT-5.22026.0473490.323625
WebJudgeVerifier=WebJudge, Backbone=o4-mini2026.0468530.341238
WebJudgeVerifier=WebJudge, Base Model=o4-mini2026.0468530.341238
WebVoy.Verifier=WebVoy., Backbone=GPT-4o2026.0455470.22556
WebVoyagerVerifier=WebVoy., Base Model=GPT-4o2026.0455470.22556