Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

Online Reinforcement Learning research benchmarks · SOTA2 Research

Online Reinforcement Learning

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

OpenAI Gym MuJoCo Normalized v4

95.5Normalized Mean Return

50

Feb 26, 2026

1.343Reward

10

Jun 24, 2026

19.4Reward

10

Jun 24, 2026

62.28Success Rate

9

Jun 8, 2026

Antmaze large-diverse

93.5Score

8

Feb 26, 2026

Antmaze large-play

94.8Score

8

Feb 26, 2026

Antmaze medium-diverse

98.5Score

8

Feb 26, 2026

Antmaze medium-play

98.7Score

8

Feb 26, 2026

Antmaze umaze-diverse

99.9Score

8

Feb 26, 2026

99.9Score

8

Feb 26, 2026

MinAtar (|A|=216, k=3) Macro-Action (online)

11.98Breakout Score

5

May 13, 2026

WalkerWalk DMControl (final)

919.61Normalized Return

5

Feb 26, 2026

HopperStand DMControl (final)

874.63Normalized Return

5

Feb 26, 2026

FishSwim DMControl (final)

641.01Normalized Return

5

Feb 26, 2026

FingerTurnHard DMControl (final)

884.59Normalized Return

5

Feb 26, 2026

DMControl FingerSpin (final)

903.92Normalized Return

5

Feb 26, 2026

CheetahRun DMControl (final)

902.24Normalized Return

5

Feb 26, 2026

Popular datasets

Antmaze AndroidWorld Clevr DMControl Meta-World

Follow for updates

Join SOTA2 Contact