List your company

Off-policy evaluation on HotpotQA (DPO)

0.9RhoADWM
-0.0360.2070.450.693Jun 4, 2026

Evaluation results

MethodLinks
ADWMn=52026.060.90.110.037
DRn=52026.060.9——
ISn=52026.060.9——
WISn=52026.060.9——
FQEn=52026.060.1——
DMn=52026.060——