Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Human Preference Alignment on HPDv2

0.3735HPS-v2.1

TreeGRPO

0.268460.295730.3230.35027Dec 9, 2025
Updated 4d ago

Evaluation Results

MethodLinks
2025.12
0.37351.32946.50940.3703
2025.12
0.36491.22636.42950.3612
2025.12
0.35561.36686.3080.3769
2025.12
0.27581.00675.94580.39
2025.12
0.27250.8875.95190.3996