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

DeepMind Control Suite dataset and benchmarks – Page 4 · SOTA2 Research

Research Datasets

DeepMind Control Suite

Benchmarks

Task NameDataset NameSOTA MethodSort

Task NameDataset NameSOTA ResultTrendResults

Reinforcement Learning

DeepMind Control Suite cheetah_run medium

479Episode Return

6

Reinforcement Learning

DeepMind Control Suite cheetah run easy

508Episode Return

6

Reinforcement Learning

DeepMind Control Suite walker_walk, hard

842Episode Return

6

Reinforcement Learning

DeepMind Control Suite walker_walk medium

902Episode Return

6

Reinforcement Learning

DeepMind Control Suite walker_walk easy

884Episode Return

6

Visual Reinforcement Learning

DeepMind Control Suite 500K steps

408Quadruped Walk Score

6

Offline Reinforcement Learning

DeepMind Control Suite walker-walk plan2explore

81.9Mean Normalized Return

6

Offline Reinforcement Learning

DeepMind Control Suite hopper-stand plan2explore

54.1Mean Normalized Return

6

Offline Reinforcement Learning

DeepMind Control Suite hopper-stand medium-replay

63.1Mean Normalized Return

6

Offline Reinforcement Learning

DeepMind Control Suite hopper-stand medium

65Mean Normalized Return

6

Offline Reinforcement Learning

DeepMind Control Suite hopper-stand expert

72.1Mean Normalized Return

6

Offline Reinforcement Learning

DeepMind Control Suite cheetah-run plan2explore

28.5Mean Normalized Return

6

Offline Reinforcement Learning

DeepMind Control Suite cheetah-run medium

62.6Mean Normalized Return

6

Physical state prediction

DeepMind Control Suite Panda Easy tasks (random policy)

0.001MSE

6

Continuous Control

DeepMind Control Suite Walker Walk

950.7Reward

6

Imitation Learning

DeepMind Control Suite image-based

0.94Cartpole Score

6

DeepMind Control Suite (DMC)

158AUC (Hopper Hop)

5

DeepMind Control Suite (DMC)

508AUC (Walker Run)

5

DeepMind Control Suite (DMC)

776Walker Walk AUC

5

DeepMind Control Suite (DMC)

698,200Steps to 80% Return

5

DeepMind Control Suite (DMC)

900.8Steps (k)

5

DeepMind Control Suite (DMC)

165.5Steps to 80% Return (k)

5

Cartpole Swingup Sparse

DeepMind Control Suite (DMC)

780.9Steps (k)

5

Multi-task Control Performance and Consistency

DeepMind Control Suite Average

117.9Average AUC

5

Cartpole Swingup

DeepMind Control Suite Cartpole Swingup

65.8Final Return (100k steps)

5

Page 4 of 7

Popular tasks

Continuous Control Reinforcement Learning Offline Reinforcement Learning Locomotion Cheetah Run

Follow for updates

Join SOTA2 Contact