Atari Atlantis Environment

Overview

The player controls the last defenses of the City of Atlantis against the Gorgon invaders. The city has seven bases, which are vulnerable to attack. Three of these have firepower capabilities to destroy the Gorgon ships before they manage to fire death rays at one of the settlements. The gun bases have fixed cannons; the center base fires straight up, while the far left and far right bases fire diagonally upwards across the screen. The center cannon also creates a shield that protects the settlements from the death rays, so once the center cannon is destroyed, the remaining settlements become vulnerable to attack. The enemy ships pass back and forth from left to right four times before they enter firing range, giving an ample opportunity to blow them away. Lost bases can be regained by destroying enough Gorgon ships.

Description from Wikipedia

Performances of RL Agents

We list various reinforcement learning algorithms that were tested in this environment. These results are from RL Database. If this page was helpful, please consider giving a star!

Star

Human Starts

Result Algorithm Source
911091.0 A3C FF Asynchronous Methods for Deep Reinforcement Learning
875822.0 A3C LSTM Asynchronous Methods for Deep Reinforcement Learning
814684.0 Rainbow Rainbow: Combining Improvements in Deep Reinforcement Learning
772392.0 A3C FF 1 day Asynchronous Methods for Deep Reinforcement Learning
629166.5 Gorila DQN Massively Parallel Methods for Deep Reinforcement Learning
593642.0 Prioritized DDQN (prop, tuned) Prioritized Experience Replay
445360.0 DuDQN Dueling Network Architectures for Deep Reinforcement Learning
423252.0 PDD DQN Dueling Network Architectures for Deep Reinforcement Learning
330647.0 Prioritized DDQN (rank, tuned) Prioritized Experience Replay
319688.0 DDQN (tuned) Deep Reinforcement Learning with Double Q-learning
289803.0 Distributional DQN Rainbow: Combining Improvements in Deep Reinforcement Learning
260556.0 DDQN Deep Reinforcement Learning with Double Q-learning
207526.0 Prioritized DQN (rank) Prioritized Experience Replay
76108.0 DQN Massively Parallel Methods for Deep Reinforcement Learning
26575.0 Human Massively Parallel Methods for Deep Reinforcement Learning
13463.0 Random Massively Parallel Methods for Deep Reinforcement Learning

No-op Starts

Result Algorithm Source
3433182.0 ACKTR Scalable trust-region method for deep reinforcement learning using Kronecker-factored approximation
1046625 QR-DQN-0 Distributional Reinforcement Learning with Quantile Regression
978200 IQN Implicit Quantile Networks for Distributional Reinforcement Learning
972175 NoisyNet DuDQN Noisy Networks for Exploration
971850 QR-DQN-1 Distributional Reinforcement Learning with Quantile Regression
968179.5 Reactor ND The Reactor: A fast and sample-efficient Actor-Critic agent for Reinforcement Learning
923733 NoisyNet DQN Noisy Networks for Exploration
902742 DuDQN Noisy Networks for Exploration
876000 DQN Noisy Networks for Exploration
849967.5 IMPALA (deep) IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures
841075 C51 A Distributional Perspective on Reinforcement Learning
826659.5 Rainbow Rainbow: Combining Improvements in Deep Reinforcement Learning
773355.5 IMPALA (shallow) IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures
465700 NoisyNet A3C Noisy Networks for Exploration
460430.5 IMPALA (deep, multitask) IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures
422700 A3C Noisy Networks for Exploration
395762.0 PDD DQN Dueling Network Architectures for Deep Reinforcement Learning
382572.0 DuDQN Dueling Network Architectures for Deep Reinforcement Learning
308258 Reactor The Reactor: A fast and sample-efficient Actor-Critic agent for Reinforcement Learning
302831.0 Reactor The Reactor: A fast and sample-efficient Actor-Critic agent for Reinforcement Learning
279987.0 DQN A Distributional Perspective on Reinforcement Learning
273895.0 Distributional DQN Rainbow: Combining Improvements in Deep Reinforcement Learning
106056.0 DDQN A Distributional Perspective on Reinforcement Learning
100069.16 Gorila DQN Massively Parallel Methods for Deep Reinforcement Learning
85641 DQN Human-level control through deep reinforcement learning
64758.0 DDQN Deep Reinforcement Learning with Double Q-learning
62687 Linear Human-level control through deep reinforcement learning
29028.1 Human Dueling Network Architectures for Deep Reinforcement Learning
29028.1 Human Human-level control through deep reinforcement learning
12850.0 Random Human-level control through deep reinforcement learning
852.9 Contingency Human-level control through deep reinforcement learning

Normal Starts

Result Algorithm Source
2311815.0 PPO Proximal Policy Optimization Algorithm
1841376.0 ACER Proximal Policy Optimization Algorithm
729265.3 A2C Proximal Policy Optimization Algorithm