List your company

ResearchFormerly WizWand

Breakout

Benchmarks

TaskDatasetSOTA resultTrendResults

Adversarial Attack

Breakout Black-box discrete (test)

130.26Cumulative Reward

36

Adversarial Attack

Breakout White-box discrete (test)

65.73Cumulative Reward

36

Single-agent Reinforcement Learning (SARL)

Breakout Poisoned

570.9Average Episode Return

9

Single-agent Reinforcement Learning (SARL)

Breakout Clean

505.5Average Episode Return

9

Backdoor Attack on Reinforcement Learning

Breakout Discrete (evaluation)

495.2Baseline Reward

5

Reinforcement Learning

Breakout Atari 2600 (test)

168Average Total Reward

5

Reinforcement Learning

Breakout

9.8Return

4

Deep Reinforcement Learning

Breakout Atari 2600

364IQM Return

4

Video Game Playing

Breakout Atari Gymnasium

100ASR

3