Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Offline Reinforcement Learning on Maze2D medium v1

49.3Normalized Return

PrivORL-j

3.74815.57427.439.226Dec 8, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.12
49.3
2025.12
38
2025.12
35
2025.12
32.8
2025.12
31.5
2025.12
31.1
2025.12
26.6
2025.12
23.4
2025.12
19
2025.12
18.1
2025.12
16.4
2025.12
14.1
2025.12
13.5
2025.12
10.8
2025.12
10.2
2025.12
7.6
2025.12
6.7
2025.12
5.5