List your company
ResearchFormerly WizWand

Gradient Estimation on Fitted Q-evaluation

0.0079RMSEOracle
0.0047680.0259090.047050.068191May 20, 2026
Updated 5mo ago

Evaluation results

MethodLinks
OracleN=32002026.050.0079
OBiGradN=32002026.050.008
PIN=32002026.050.0081
OBiGradN=16002026.050.0112
OracleN=16002026.050.0112
PIN=16002026.050.0124
OracleN=8002026.050.0157
OBiGradN=8002026.050.0167
PIN=8002026.050.0208
OracleN=4002026.050.0216
OBiGradN=4002026.050.0262
OracleN=2002026.050.0314
OBiGradN=2002026.050.0486
PIN=4002026.050.0511
PIN=2002026.050.0862