Offline-to-Online Reinforcement Learning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
100Success Rate (Cube-Double, Task 2)
6
May 11, 2026
100Lift Success Rate
6
May 11, 2026
650.5Regret
4
Feb 26, 2026
131.4Regret
4
Feb 26, 2026
293.7Regret
4
May 15, 2026
49.37Return (HalfCheetah Random)
3
May 15, 2026
12.43Walker2d Return (Random)
2
May 15, 2026