List your company

Offline Reinforcement Learning on Maze2D large v1

220.66Normalized ReturnInverter
-6.47652.492111.46170.428Dec 8, 2025Jan 4, 2026Feb 1, 2026Feb 28, 2026Mar 28, 2026Apr 24, 2026May 22, 2026

Evaluation results

MethodLinks
InverterK=16, #NN-passes=502026.05220.660.21.5914.393.71
AWAC#NN-passes=8002026.05209.138.191.28—1.0210.9
SAC-N#NN-passes=8002026.05204.761.191.8—1.4415.4
Diffuser#NN-passes=2562026.05123.0762.2214.32173.6839.3
TD3+BC#NN-passes=8002026.0597.125.411.67—1.3414.3
EDAC#NN-passes=8002026.0595.622.921.74—1.3914.9
ReBRAC#NN-passes=8002026.0578.3361.770.21—1681.8
IQL#NN-passes=8002026.0561.723.51.69—1.3614.5
PrivORL-jRL Algorithm=IQL, epsilon=102025.1237.7—————
DT#NN-passes=8002026.0535.6628.23.57—2.8630.5
PrivORL-jRL Algorithm=TD3PLUSBC, epsilon=102025.1235.6—————
PrivORL-j-URL Algorithm=IQL, epsilon=102025.1231.9—————
PrivORL-jRL Algorithm=EDAC, epsilon=102025.1230.5—————
PrivORL-jRL Algorithm=TD3PLUSBC, epsilon=12025.1229.5—————
DP-TransformerRL Algorithm=TD3PLUSBC, epsilon=102025.1228.3—————
CQL#NN-passes=8002026.0523.7536.71.84—1.4715.7
DP-TransformerRL Algorithm=IQL, epsilon=102025.1223.4—————
DP-TransformerRL Algorithm=EDAC, epsilon=102025.1221.8—————
PrivORL-jRL Algorithm=EDAC, epsilon=12025.1220.5—————
PrivORL-j-URL Algorithm=EDAC, epsilon=102025.1219.8—————
PrivORL-j-URL Algorithm=TD3PLUSBC, epsilon=102025.1214.3—————
BC-10%#NN-passes=8002026.0511.325.11.25—999.710.7
PrivORL-j-URL Algorithm=EDAC, epsilon=12025.1210.9—————
PrivORL-jRL Algorithm=IQL, epsilon=12025.129.9—————
PrivORL-j-URL Algorithm=TD3PLUSBC, epsilon=12025.129.9—————
PrivORL-j-URL Algorithm=IQL, epsilon=12025.129.8—————
DP-TransformerRL Algorithm=EDAC, epsilon=12025.127.6—————
DP-TransformerRL Algorithm=TD3PLUSBC, epsilon=12025.126.3—————
DP-TransformerRL Algorithm=IQL, epsilon=12025.124.9—————
BC#NN-passes=8002026.052.264.391.7—1.3614.6