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

Reinforcement Learning research benchmarks – Page 24 · SOTA2 Research

Reinforcement Learning

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

D4RL Walker Random

Adaptive Policy Selection and Fine-Tuning

47.1Mean Normalized Return

5

May 7, 2026

D4RL Cheetah Medium-Expert

98.7Mean Normalized Return

5

May 7, 2026

D4RL Cheetah Medium

Adaptive Policy Selection and Fine-Tuning

92Mean Normalized Return

5

May 7, 2026

D4RL Cheetah Random

91.9Mean Normalized Return

5

May 7, 2026

D4RL Hopper Medium-Expert

Adaptive Policy Selection and Fine-Tuning

77.2Mean Normalized Return

5

May 7, 2026

Atari Learning Environment Atari-5

115Battle Zone Score

5

May 6, 2026

426.9Score

5

Apr 6, 2026

4,595.3Max Return

5

Mar 31, 2026

Inverted Pendulum

1,000Maximum Evaluation Return

5

Mar 31, 2026

6.48Maximum Return

5

Mar 31, 2026

Safety Gym (test)

19.13Point Goal 1 Success

5

Mar 31, 2026

1,005Return

5

Feb 26, 2026

24.1Return

5

Feb 26, 2026

23Return

5

Feb 26, 2026

obstacle4 non-convex

33.4Return

5

Feb 26, 2026

obstacle3 non-convex

33.4Return

5

Feb 26, 2026

34.2Return

5

Feb 26, 2026

32.9Return

5

Feb 26, 2026

91.3Return

5

Feb 26, 2026

171Return

5

Feb 26, 2026

200Return

5

Feb 26, 2026

3.28Avg Performance Gain

5

Feb 26, 2026

MuJoCo Ant epsilon=0.15 (test)

5,381Natural Return

5

Feb 26, 2026

MuJoCo Halfcheetah epsilon=0.15 (test)

5,048Natural Return

5

Feb 26, 2026

MuJoCo Walker2d epsilon=0.05 (test)

4,875Natural Return

5

Feb 26, 2026

Page 24 of 41

Popular datasets

Atari Atari 2600 halfcheetah DMControl hopper

Follow for updates

Join SOTA2 Contact