List your company

Pusher Environment Evaluation

100Convergence RateDAgger
16.838.46081.6Nov 29, 2025

Evaluation results

MethodLinks
DAggerInitial expert dataset size (M)=1,000, Training steps (Ttrain)=20002025.11100480,17220,000
DAggerInitial expert dataset size (M)=2,000, Training steps (Ttrain)=20002025.1110034,07520,000
DAggerInitial expert dataset size (M)=5,000, Training steps (Ttrain)=20002025.11100340,10220,000
DAggerInitial expert dataset size (M)=10,000, Training steps (Ttrain)=20002025.11100700,14120,000
DAggerInitial expert dataset size (M)=20,000, Training steps (Ttrain)=20002025.11100580,24020,000
EnsembleInitial expert dataset size (M)=1,000, Training steps (Ttrain)=20002025.11100293,2031,415,221
EnsembleInitial expert dataset size (M)=2,000, Training steps (Ttrain)=20002025.11100274,154114,782
EnsembleInitial expert dataset size (M)=5,000, Training steps (Ttrain)=20002025.11100237,126911,130
EnsembleInitial expert dataset size (M)=10,000, Training steps (Ttrain)=20002025.1110021,77495,745
EnsembleInitial expert dataset size (M)=20,000, Training steps (Ttrain)=20002025.11100469,10998,739
ThriftyInitial expert dataset size (M)=1,000, Training steps (Ttrain)=20002025.1110049,087129,456
ThriftyInitial expert dataset size (M)=2,000, Training steps (Ttrain)=20002025.11100651,154130,830
CRSAILInitial expert dataset size (M)=1,000, Training steps (Ttrain)=20002025.11100260,153708,422
CRSAILInitial expert dataset size (M)=2,000, Training steps (Ttrain)=20002025.1110022,257761,367
CRSAILInitial expert dataset size (M)=5,000, Training steps (Ttrain)=20002025.1110018,246463,258
CRSAILInitial expert dataset size (M)=10,000, Training steps (Ttrain)=20002025.1110018,467398,146
CRSAILInitial expert dataset size (M)=20,000, Training steps (Ttrain)=20002025.1110024,686495,155
ThriftyInitial expert dataset size (M)=5,000, Training steps (Ttrain)=20002025.1180854,162128,381
ThriftyInitial expert dataset size (M)=10,000, Training steps (Ttrain)=20002025.1140109,862128,048
ThriftyInitial expert dataset size (M)=20,000, Training steps (Ttrain)=20002025.112013,270129,033