Atari Games

The Atari 2600 Games task (and dataset) involves training an agent to achieve high game scores.

( Image credit: Playing Atari with Deep Reinforcement Learning )

Papers

Recently Added Most Hyped Most Active Needs Verification Most Verified

Showing 351–375 of 625 papers

Title	Date	Tasks	Status
Convex Regularization in Monte-Carlo Tree Search	Jul 1, 2020	Atari GamesDecision Making	—Unverified
Reinforcement Learning and its Connections with Neuroscience and Psychology	Jun 25, 2020	Atari GamesDecision Making	—Unverified
RL Unplugged: A Suite of Benchmarks for Offline Reinforcement Learning	Jun 24, 2020	Atari GamesDQN Replay Dataset	CodeCode Available
On Effective Parallelization of Monte Carlo Tree Search	Jun 15, 2020	Atari Games	—Unverified
A Decentralized Policy Gradient Approach to Multi-task Reinforcement Learning	Jun 8, 2020	Atari GamesMulti-Task Learning	—Unverified
The Value-Improvement Path: Towards Better Representations for Reinforcement Learning	Jun 3, 2020	Atari Gamesreinforcement-learning	—Unverified
Gradient Monitored Reinforcement Learning	May 25, 2020	Atari Gamescontinuous-control	—Unverified
A Metric Learning Approach to Anomaly Detection in Video Games	May 20, 2020	Anomaly DetectionAtari Games	CodeCode Available
Local and Global Explanations of Agent Behavior: Integrating Strategy Summaries with Saliency Maps	May 18, 2020	Atari GamesDecision Making	CodeCode Available
A Simple Imitation Learning Method via Contrastive Regularization	May 16, 2020	Atari GamesImitation Learning	—Unverified
Accelerating Deep Neuroevolution on Distributed FPGAs for Reinforcement Learning Problems	May 10, 2020	Atari GamesComputational Efficiency	—Unverified
Model Based Reinforcement Learning for Atari	May 1, 2020	Atari Gamesmodel	—Unverified
Explain Your Move: Understanding Agent Actions Using Focused Feature Saliency	May 1, 2020	Atari GamesBoard Games	CodeCode Available
Episodic Reinforcement Learning with Associative Memory	May 1, 2020	Atari GamesDeep Reinforcement Learning	—Unverified
Learning Dialog Policies from Weak Demonstrations	Apr 23, 2020	Atari GamesDeep Reinforcement Learning	—Unverified
Self Punishment and Reward Backfill for Deep Q-Learning	Apr 10, 2020	Atari GamesDeep Reinforcement Learning	CodeCode Available
Using Generative Adversarial Nets on Atari Games for Feature Extraction in Deep Reinforcement Learning	Apr 6, 2020	Atari GamesDeep Reinforcement Learning	—Unverified
Exploring Unknown States with Action Balance	Mar 10, 2020	Atari GamesDeep Reinforcement Learning	CodeCode Available
Relevance-Guided Modeling of Object Dynamics for Reinforcement Learning	Mar 3, 2020	Atari GamesDeep Reinforcement Learning	—Unverified
A Self-Tuning Actor-Critic Algorithm	Feb 28, 2020	Atari Gamesreinforcement-learning	—Unverified
On Catastrophic Interference in Atari 2600 Games	Feb 28, 2020	Atari GamesDeep Reinforcement Learning	CodeCode Available
Efficiently Guiding Imitation Learning Agents with Human Gaze	Feb 28, 2020	Atari GamesImitation Learning	—Unverified
ConQUR: Mitigating Delusional Bias in Deep Q-learning	Feb 27, 2020	Atari GamesQ-Learning	CodeCode Available
Disentangling Controllable Object through Video Prediction Improves Visual Reinforcement Learning	Feb 21, 2020	Atari GamesObject	—Unverified
Value-driven Hindsight Modelling	Feb 19, 2020	Atari GamesReinforcement Learning	—Unverified

Show:10 25 50

← PrevPage 15 of 25Next →

All datasets Atari 2600 Breakout Atari 2600 Freeway Atari 2600 Q*Bert Atari 2600 Seaquest Atari 2600 Venture Atari 2600 Space Invaders Atari 2600 Frostbite Atari 2600 Gravitar Atari 2600 Pong Atari 2600 Private Eye Atari 2600 Montezuma's Revenge Atari 2600 Alien

Benchmark Results

#	Model	Metric	Claimed	Verified	Status
1	GDI-H3(200M frames)	Score	864	—	Unverified
2	GDI-I3(200M frames)	Score	864	—	Unverified
3	GDI-I3	Score	864	—	Unverified
4	GDI-H3	Score	864	—	Unverified
5	Bootstrapped DQN	Score	855	—	Unverified
6	FQF	Score	854.2	—	Unverified
7	R2D2	Score	837.7	—	Unverified
8	Ape-X	Score	800.9	—	Unverified
9	Agent57	Score	790.4	—	Unverified
10	IMPALA (deep)	Score	787.34	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	GDI-I3	Score	34	—	Unverified
2	Go-Explore	Score	34	—	Unverified
3	IQN	Score	34	—	Unverified
4	QR-DQN-1	Score	34	—	Unverified
5	GDI-H3	Score	34	—	Unverified
6	GDI-H3(200M frames)	Score	34	—	Unverified
7	TRPO-hash	Score	34	—	Unverified
8	NoisyNet-Dueling	Score	34	—	Unverified
9	C51 noop	Score	33.9	—	Unverified
10	Bootstrapped DQN	Score	33.9	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	Agent57	Score	580,328.14	—	Unverified
2	QR-DQN-1	Score	572,510	—	Unverified
3	R2D2	Score	408,850	—	Unverified
4	IMPALA (deep)	Score	351,200.12	—	Unverified
5	Ape-X	Score	302,391.3	—	Unverified
6	A2C + SIL	Score	104,975.6	—	Unverified
7	MuZero (Res2 Adam)	Score	94,906.25	—	Unverified
8	DreamerV2	Score	94,688	—	Unverified
9	MuZero	Score	72,276	—	Unverified
10	DNA	Score	52,398	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	GDI-H3	Score	1,000,000	—	Unverified
2	GDI-H3(200M frames)	Score	1,000,000	—	Unverified
3	Agent57	Score	999,997.63	—	Unverified
4	R2D2	Score	999,996.7	—	Unverified
5	MuZero	Score	999,976.52	—	Unverified
6	MuZero (Res2 Adam)	Score	999,659.18	—	Unverified
7	GDI-I3	Score	943,910	—	Unverified
8	Ape-X	Score	392,952.3	—	Unverified
9	C51 noop	Score	266,434	—	Unverified
10	Duel noop	Score	50,254.2	—	Unverified