List your company
ResearchFormerly WizWand

Reinforcement Learning on Hopper v4

27,721,263Average Returnpop-SAN
-1,107,318.66,377,024.713,861,36821,345,711.3Jan 29, 2026Feb 22, 2026Mar 18, 2026Apr 11, 2026May 5, 2026May 29, 2026Jun 22, 2026
Updated 4mo ago

Evaluation results

MethodLinks
pop-SAN2026.0227,721,263
MDC-SAN2026.023,446,131
ANNBase Algorithm=TD32026.023,410,164
ILC-SAN2026.023,403,148
PT-LIF2026.023,385,157
ANN-SNN2026.023,098,281
DSN2026.02356,568
Vanilla LIF2026.02352,094
S-PLIFFrame=Vanilla2026.013,462
PLIFFrame=Vanilla2026.013,414
PLIFFrame=PT2026.013,384
S-PLIFFrame=PT2026.013,380
C-DSACNumber of runs=100, Selection=Best models in training runs2026.043,352
ReLUFrame=ANN2026.013,349
PDAEnvironment steps=1M, Number of seeds=10, Tests per epoch=102026.032,944.3
FedNPGNumber of Agents=82026.052,736
FedNPG-ADMMNumber of Agents=82026.052,719
Fixed βNumber of seeds=5, Evaluation window=last 10% of training2026.062,616
PPO-ClipNumber of seeds=5, Evaluation window=last 10% of training2026.062,598
per-sample PPO-KLNumber of seeds=5, Evaluation window=last 10% of training2026.062,598
FedNPG-ADMMNumber of Agents=42026.052,507
FedNPGNumber of Agents=22026.052,468
FedNPGNumber of Agents=42026.052,458
FedNPG-ADMMNumber of Agents=22026.052,384
PPOEnvironment steps=1M, Number of seeds=10, Tests per epoch=102026.032,329.7
Adaptive βNumber of seeds=5, Evaluation window=last 10% of training2026.062,236
TRPOEnvironment steps=1M, Number of seeds=10, Tests per epoch=102026.032,017
NPGEnvironment steps=1M, Number of seeds=10, Tests per epoch=102026.031,650.8
FedNPGNumber of Agents=12026.051,644
FedNPG-ADMMNumber of Agents=12026.051,473