List your company

Alignment with human judgment on game verification on VERIGAME

92.2Top-5 AccuracyGameGen-Verifier
-2.54422.05346.6571.247May 8, 2026

Evaluation results

MethodLinks
GameGen-VerifierBackend=Codex2026.0592.291.210095.4443
GameGen-VerifierBackend=Claude Code (Opus)2026.0590.289.210094.3513
GameGen-VerifierBackend=Claude Code (Kimi)2026.0586.689.396.292.6614
AaaV-CEBackend=Codex2026.0558.896.848.464.57,356
AaaV-CEBackend=Claude Code (Opus)2026.0557.693.547.5638,245
AaaV-CEBackend=Claude Code (Kimi)2026.055192.352.266.79,072
AaaV-DirectBackend=Claude Code (Kimi)2026.053.381.22.14537
AaaV-DirectBackend=Claude Code (Opus)2026.051.371.41.12.2552
AaaV-DirectBackend=Codex2026.051.166.70.91.8387