List your company

Strategic Gaming on Rock-Paper-Scissors (test)

1.6324RGRPO
0.9793841.1489171.318451.487983Jan 22, 2026

Evaluation results

MethodLinks
GRPOFramework=GameTalk2026.011.63240.6891-0.61190.8421.50560.73360.1840.0824
DPO (pairs)Framework=GameTalk2026.011.56340.6288-0.55320.87721.54470.70760.17910.1132
STaRFramework=GameTalk2026.011.37880.4864-0.50380.78291.72540.51130.39750.0913
UntrainedFramework=GameTalk2026.011.00450.0483-0.56290.52251.58360.31250.40630.2813