HyperAI超神経
ホームニュース最新論文チュートリアルデータセット百科事典SOTALLMモデルGPU ランキング学会
検索
サイトについて
HyperAI超神経

Atari Games On Atari 2600 Tutankham

評価指標

Score

評価結果

このベンチマークにおける各モデルのパフォーマンス結果

モデル名
Score
Paper TitleRepository
DQN hs45.6Deep Reinforcement Learning with Double Q-learning
SARSA98.2--
DDQN (tuned) noop218.4Dueling Network Architectures for Deep Reinforcement Learning
Prior noop204.6Prioritized Experience Replay
GDI-I3423.9Generalized Data Distribution Iteration-
R2D2395.3Recurrent Experience Replay in Distributed Reinforcement Learning-
MuZero491.48Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model
Prior+Duel noop245.9Dueling Network Architectures for Deep Reinforcement Learning
NoisyNet-Dueling269Noisy Networks for Exploration
CGP0Evolving simple programs for playing Atari games
DDQN+Pop-Art noop183.9Learning values across many orders of magnitude-
GDI-I3423.9GDI: Rethinking What Makes Reinforcement Learning Different From Supervised Learning-
DDQN (tuned) hs92.2Deep Reinforcement Learning with Double Q-learning
A3C LSTM hs144.2Asynchronous Methods for Deep Reinforcement Learning
GDI-H3418.2Generalized Data Distribution Iteration-
C51 noop280.0A Distributional Perspective on Reinforcement Learning
Advantage Learning245.22Increasing the Action Gap: New Operators for Reinforcement Learning
DQN noop68.1Deep Reinforcement Learning with Double Q-learning
Best Learner114.3The Arcade Learning Environment: An Evaluation Platform for General Agents
ASL DDQN252.9Train a Real-world Local Path Planner in One Hour via Partially Decoupled Reinforcement Learning and Vectorized Diversity-
0 of 44 row(s) selected.