List your company

Agreement with process human labels on CUAVerifierBench (Internal Dataset (n=140))

81AccuracyUniversal Verifier (UV)
60.265.67176.4Apr 5, 2026

Evaluation results

MethodLinks
Universal Verifier (UV)Verifier=UV, Backbone=GPT-5.22026.0481860.59244
Universal VerifierVerifier=UV, Base Model=GPT-5.22026.0481860.59244
WebJudgeVerifier=WebJudge, Backbone=o4-mini2026.0466700.324025
WebJudgeVerifier=WebJudge, Base Model=o4-mini2026.0466700.324025
WebVoy.Verifier=WebVoy., Backbone=GPT-5.22026.0464650.344910
WebVoy.Verifier=WebVoy., Backbone=GPT-4o2026.0462700.173152
WebVoyagerVerifier=WebVoy., Base Model=GPT-4o2026.0462700.173152
WebJudgeVerifier=WebJudge, Backbone=GPT-5.22026.0461570.3594