List your company
ResearchFormerly WizWand

Open-loop Planning on Bench2Drive

0.38Average L2 ErrorDriveMOE
0.24961.12982.012.8902May 26, 2025Jul 16, 2025Sep 6, 2025Oct 28, 2025Dec 19, 2025Feb 9, 2026Apr 2, 2026
Updated 6mo ago

Evaluation results

MethodLinks
DriveMOETraining Expert=Think2Drive2026.040.38
ReasonPlan2025.050.61
DriveTransformer2025.110.62
DriveTransformerTraining Expert=Think2Drive2026.040.62
DriveTransformer2025.080.62
CogDriver-Agent2025.080.63
SeerDriveTraining Expert=Think2Drive2026.040.66
OrionTraining Expert=Think2Drive2026.040.68
DiFSD2025.110.7
BridgeAD2025.110.71
UniDriveVLATraining Expert=Think2Drive2026.040.72
UniAD-BaseModel size=Base2025.050.73
UniADModel variant=Base2025.110.73
UniADTraining Expert=Think2Drive2026.040.73
UniADModel Size=Base2025.080.73
WPT-Student2025.110.75
WPT-Teacher2025.110.76
Baseline2025.110.79
UniAD-TinyModel size=Tiny2025.050.8
UniADModel variant=Tiny2025.110.8
UniADModel Size=Tiny2025.080.8
MomAD2025.050.82
MomADTraining Expert=Think2Drive2026.040.82
ReAL-AD2025.080.84
SparseDrive2025.110.87
SparseDriveTraining Expert=Think2Drive2026.040.87
MomAD2025.080.87
VAD2025.050.91
VADModel variant=Base2025.110.91
VADTraining Expert=Think2Drive2026.040.91
VAD2025.080.91
ThinkTwiceExpert feature distillation=true2025.050.95
ThinkTwiceexpert feature distillation=true2025.110.95
ThinkTwiceTraining Expert=Think2Drive2026.040.95
ThinkTwice2025.080.95
DriveAdapterExpert feature distillation=true2025.051.01
DriveAdapterexpert feature distillation=true2025.111.01
DriveAdapterTraining Expert=Think2Drive2026.041.01
DriveAdapter2025.081.01
VADModel variant=Tiny2025.111.15
TCPExpert feature distillation=true2025.051.7
TCP-trajexpert feature distillation=true2025.111.7
TCP2025.081.7
AD-MLP2025.053.64
AD-MLP2025.113.64
AD-MLPTraining Expert=Think2Drive2026.043.64
AD-MLP2025.083.64