List your company

Human Agreement Evaluation on Interactive Agent Evaluation Criteria

0.7Average Agreement ScoreOnline Agent-as-a-Judge
0.31520.41510.5150.6149Jun 6, 2026

Evaluation results

MethodLinks
Online Agent-as-a-Judge2026.060.70.580.680.330.760.820.720.730.96
Offline Agent-as-a-Judge2026.060.40.470.380.140.520.550.440.330.17
Offline LLM-as-a-Judge2026.060.330.420.3800.260.450.440.330.17