Atari Games On Atari 2600 Double Dunk

평가 지표

Score

평가 결과

이 벤치마크에서 각 모델의 성능 결과

비교 표

모델 이름	Score
distributional-reinforcement-learning-with-1	21.9
mastering-atari-go-chess-and-shogi-by	23.94
self-imitation-learning	21.5
dueling-network-architectures-for-deep	0.1
massively-parallel-methods-for-deep	-11.3
implicit-quantile-networks-for-distributional	5.6
dueling-network-architectures-for-deep	-12.5
deep-reinforcement-learning-with-double-q	-0.3
impala-scalable-distributed-deep-rl-with	-0.33
generalized-data-distribution-iteration	24
prioritized-experience-replay	18.5
mastering-atari-with-discrete-world-models-1	17
deep-reinforcement-learning-with-double-q	-10.7
deep-reinforcement-learning-with-double-q	-6.6
the-arcade-learning-environment-an-evaluation	24
learning-values-across-many-orders-of	-11.5
asynchronous-methods-for-deep-reinforcement	0.1
human-level-control-through-deep	-18.1
모델 19	-16.0
dueling-network-architectures-for-deep	-5.5
distributed-prioritized-experience-replay	23.5
the-reactor-a-fast-and-sample-efficient-actor	23.0
dueling-network-architectures-for-deep	-0.8
deep-reinforcement-learning-with-double-q	-6.0
deep-exploration-via-bootstrapped-dqn	3
evolving-simple-programs-for-playing-atari	2
online-and-offline-reinforcement-learning-by	23.91
policy-optimization-with-penalized-point	-7.89
the-arcade-learning-environment-an-evaluation	-13.1
gdi-rethinking-what-makes-reinforcement	24
asynchronous-methods-for-deep-reinforcement	-0.1
dna-proximal-policy-optimization-with-a-dual	-1.3
train-a-real-world-local-path-planner-in-one	0.1
increasing-the-action-gap-new-operators-for	-2.51
asynchronous-methods-for-deep-reinforcement	0.1
prioritized-experience-replay	16.0
increasing-the-action-gap-new-operators-for	-0.15
generalized-data-distribution-iteration	24
agent57-outperforming-the-atari-human	23.93
recurrent-experience-replay-in-distributed	23.7
noisy-networks-for-exploration	1
a-distributional-perspective-on-reinforcement	2.5
evolution-strategies-as-a-scalable	0.2