List your company

ResearchFormerly WizWand

OpenAI Gym

Benchmarks

TaskDatasetSOTA resultTrendResults

Continuous Control

OpenAI Gym Reacher (held-out λc ∈ {2, 3, 4})

-6.093Reward

8

2D bipedal locomotion

Hardcore (OpenAI Gym) (test)

140.1Average Return

6

2D bipedal locomotion

Basic (OpenAI Gym) (test)

312Average Return

6

Reinforcement Learning

OpenAI Gym Humanoid v1

6,849Mean Reward

3

Reinforcement Learning

OpenAI Gym Ant v1

5,007Mean Reward

3

Reinforcement Learning

OpenAI Gym Walker v1

5,867Mean Reward

3

Reinforcement Learning

OpenAI Gym Cheetah v1

6,620Mean Reward

3

Reinforcement Learning

OpenAI Gym Hopper v1

3,810Mean Reward

3

Reinforcement Learning

OpenAI Gym Swimmer v1

366Mean Reward

3

Reinforcement Learning

OpenAI Gym Humanoid

96,720Episodes

3

Reinforcement Learning

OpenAI Gym Ant

73,500Number of Episodes

3

Reinforcement Learning

OpenAI Gym Walker

14,250Episode Count

3

Reinforcement Learning

OpenAI Gym Cheetah

11,250Number of Episodes

3

Reinforcement Learning

OpenAI Gym Hopper

13,920Episode Count

3

Reinforcement Learning

OpenAI Gym Swimmer

1,550Number of Episodes

2

OpenAI Gym LunarLander

-1.871T-statistic

1

OpenAI Gym MountainCar

-6.243T-statistic

1

OpenAI Gym CarRacing

-6.399T-statistic

1

OpenAI Gym BipedalWalker

-2.064T-statistic

1

Reinforcement Learning

OpenAI Gym Walker

—Primary metric

0

Reinforcement Learning

OpenAI Gym Hopper

—Primary metric

0

Reinforcement Learning

OpenAI Gym HalfCheetah

—Primary metric

0