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

Antmaze dataset and benchmarks – Page 6 · SOTA2 Research

Research Datasets

Antmaze

Benchmarks

Task NameDataset NameSOTA MethodSort

Task NameDataset NameSOTA ResultTrendResults

Offline Reinforcement Learning

Antmaze

87.5Success Rate (umaze)

5

Imitation Learning

antmaze umaze v2

44.44Normalized return

5

AntMaze 19

22.8Success Rate (Large Maze, 7 Tasks, Undirected)

5

Goal-conditioned Reinforcement Learning

antmaze explore teleport

45Success Rate

5

Goal-conditioned Reinforcement Learning

antmaze explore large

16Success Rate

5

Goal-conditioned Reinforcement Learning

antmaze explore medium

49Success Rate

5

Goal-conditioned Reinforcement Learning

antmaze stitch teleport

60Success Rate

5

Goal-conditioned Reinforcement Learning

antmaze stitch giant

61Success Rate

5

Goal-conditioned Reinforcement Learning

antmaze navigate teleport

0.55Success Rate

5

Goal-conditioned Reinforcement Learning

antmaze navigate giant

0.69Success Rate

5

Goal-conditioned Reinforcement Learning

antmaze navigate large

90Success Rate

5

Goal-conditioned Reinforcement Learning

antmaze navigate medium

96Success Rate

5

Offline goal-conditioned RL

Visual AntMaze navigate

45.7Normalized Score

5

Reinforcement Learning

AntMaze large-play v2

89.4Final Score (1M Online Steps)

5

Reinforcement Learning

AntMaze medium-play v2

97.4Final Score (1M Online Steps)

5

Reinforcement Learning

AntMaze umaze v2

99.2Final Score

5

Offline Reinforcement Learning

AntMaze umaze

95.7Normalized Average Return

4

AntMaze

44.8Success Rate (Large Play v2)

4

Long-horizon sparse reward navigation

AntMaze Large

2,224Cumulative Episodic Rewards

4

Long-horizon sparse reward navigation

AntMaze Medium

2,299Cumulative Episodic Rewards

4

Reinforcement Learning

AntMaze Medium-Play

71.8D4RL Score

4

Goal-Conditioned Trajectory Planning

AntMaze Medium

70Success Rate

4

AntMaze Random Cells v2 (large-diverse)

58.3Success Rate

4

AntMaze Random Cells medium-diverse v2

79.3Success Rate

4

AntMaze Large-Diverse v0

47.5Success Rate

4

Page 6 of 7

Popular tasks

Navigation Offline Reinforcement Learning Reinforcement Learning Goal-conditioned Reinforcement Learning Goal Reaching

Follow for updates

Join SOTA2 Contact