Atari Games On Atari 2600 Boxing

评估指标

Score

评测结果

各个模型在此基准测试上的表现结果

		Paper Title	Repository
MuZero (Res2 Adam)	100	Online and Offline Reinforcement Learning by Planning with a Learned Model
GDI-H3	100	GDI: Rethinking What Makes Reinforcement Learning Different From Supervised Learning	-
UCT	100	The Arcade Learning Environment: An Evaluation Platform for General Agents
Agent57	100	Agent57: Outperforming the Atari Human Benchmark
GDI-H3	100	Generalized Data Distribution Iteration	-
Ape-X	100	Distributed Prioritized Experience Replay
NoisyNet-Dueling	100	Noisy Networks for Exploration
MuZero	100.00	Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model
GDI-I3	100	Generalized Data Distribution Iteration	-
IMPALA (deep)	99.96	IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures
DNA	99.9	DNA: Proximal Policy Optimization with a Dual Network Architecture
QR-DQN-1	99.9	Distributional Reinforcement Learning with Quantile Regression
IQN	99.8	Implicit Quantile Networks for Distributional Reinforcement Learning
ASL DDQN	99.6	Train a Real-world Local Path Planner in One Hour via Partially Decoupled Reinforcement Learning and Vectorized Diversity
A2C + SIL	99.6	Self-Imitation Learning
Reactor 500M	99.4	The Reactor: A fast and sample-efficient Actor-Critic agent for Reinforcement Learning	-
Duel noop	99.4	Dueling Network Architectures for Deep Reinforcement Learning
DDQN+Pop-Art noop	99.3	Learning values across many orders of magnitude	-
Prior+Duel noop	98.9	Dueling Network Architectures for Deep Reinforcement Learning
R2D2	98.5	Recurrent Experience Replay in Distributed Reinforcement Learning	-

0 of 45 row(s) selected.