| Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model | ✓ Link | 143972.03 | MuZero | 2019-11-19 |
| Recurrent Experience Replay in Distributed Reinforcement Learning | ✓ Link | 108197.0 | R2D2 | 2019-05-01 |
| Generalized Data Distribution Iteration | | 97155 | GDI-H3 | 2022-06-07 |
| Agent57: Outperforming the Atari Human Benchmark | ✓ Link | 67212.67 | Agent57 | 2020-03-30 |
| Generalized Data Distribution Iteration | | 63876 | GDI-I3 | 2022-06-07 |
| Online and Offline Reinforcement Learning by Planning with a Learned Model | ✓ Link | 33292.22 | MuZero (Res2 Adam) | 2021-04-13 |
| Implicit Quantile Networks for Distributional Reinforcement Learning | ✓ Link | 29091 | IQN | 2018-06-14 |
| Distributed Prioritized Experience Replay | ✓ Link | 24559.4 | Ape-X | 2018-03-02 |
| Mastering Atari with Discrete World Models | ✓ Link | 23625 | DreamerV2 | 2020-10-05 |
| Distributional Reinforcement Learning with Quantile Regression | ✓ Link | 22012 | QR-DQN-1 | 2017-10-27 |
| IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures | ✓ Link | 19148.47 | IMPALA (deep) | 2018-02-05 |
| DNA: Proximal Policy Optimization with a Dual Network Architecture | ✓ Link | 16293 | DNA | 2022-06-20 |
| Asynchronous Methods for Deep Reinforcement Learning | ✓ Link | 14497.9 | A3C LSTM hs | 2016-02-04 |
| Train a Real-world Local Path Planner in One Hour via Partially Decoupled Reinforcement Learning and Vectorized Diversity | ✓ Link | 14372.8 | ASL DDQN | 2023-05-07 |
| Dueling Network Architectures for Deep Reinforcement Learning | ✓ Link | 11477.0 | Prior+Duel noop | 2015-11-20 |
| Noisy Networks for Exploration | ✓ Link | 11231 | NoisyNet-Dueling | 2017-06-30 |
| Deep Reinforcement Learning with Double Q-learning | ✓ Link | 10950.6 | Prior+Duel hs | 2015-09-22 |
| Dueling Network Architectures for Deep Reinforcement Learning | ✓ Link | 10950.6 | Prior+Duel hs | 2015-11-20 |
| Learning values across many orders of magnitude | | 9011.6 | DDQN+Pop-Art noop | 2016-02-24 |
| The Reactor: A fast and sample-efficient Actor-Critic agent for Reinforcement Learning | | 8323.3 | Reactor 500M | 2017-04-15 |
| Deep Exploration via Bootstrapped DQN | ✓ Link | 8047.1 | Bootstrapped DQN | 2016-02-15 |
| Prioritized Experience Replay | ✓ Link | 7672.1 | Prior noop | 2015-11-18 |
| A Distributional Perspective on Reinforcement Learning | ✓ Link | 7203.0 | C51 noop | 2017-07-21 |
| Prioritized Experience Replay | ✓ Link | 6548.9 | Prior hs | 2015-11-18 |
| Deep Reinforcement Learning with Double Q-learning | ✓ Link | 6060.8 | DDQN (tuned) hs | 2015-09-22 |
| Asynchronous Methods for Deep Reinforcement Learning | ✓ Link | 5474.9 | A3C FF hs | 2016-02-04 |
| Policy Optimization With Penalized Point Probability Distance: An Alternative To Proximal Policy Optimization | ✓ Link | 5400.13 | POP3D | 2018-07-02 |
| Dueling Network Architectures for Deep Reinforcement Learning | ✓ Link | 5393.2 | DDQN (tuned) noop | 2015-11-20 |
| Dueling Network Architectures for Deep Reinforcement Learning | ✓ Link | 4621.0 | Duel noop | 2015-11-20 |
| Deep Reinforcement Learning with Double Q-learning | ✓ Link | 4280.4 | DQN noop | 2015-09-22 |
| Dueling Network Architectures for Deep Reinforcement Learning | ✓ Link | 3994.8 | Duel hs | 2015-11-20 |
| Asynchronous Methods for Deep Reinforcement Learning | ✓ Link | 3746.1 | A3C FF (1 day) hs | 2016-02-04 |
| Increasing the Action Gap: New Operators for Reinforcement Learning | ✓ Link | 3661.51 | Advantage Learning | 2015-12-15 |
| Deep Reinforcement Learning with Double Q-learning | ✓ Link | 3489.3 | DQN hs | 2015-09-22 |
| Human level control through deep reinforcement learning | ✓ Link | 3359.0 | Nature DQN | 2015-02-25 |
| Increasing the Action Gap: New Operators for Reinforcement Learning | ✓ Link | 3304.33 | Persistent AL | 2015-12-15 |
| Self-Imitation Learning | ✓ Link | 1812 | A2C + SIL | 2018-06-14 |
| Evolution Strategies as a Scalable Alternative to Reinforcement Learning | ✓ Link | 1673.9 | ES FF (1 hour) noop | 2017-03-10 |
| The Arcade Learning Environment: An Evaluation Platform for General Agents | ✓ Link | 1512.2 | UCT | 2012-07-19 |
| Massively Parallel Methods for Deep Reinforcement Learning | ✓ Link | 1195.8 | Gorila | 2015-07-15 |
| Evolving simple programs for playing Atari games | ✓ Link | 890.4 | CGP | 2018-06-14 |
| The Arcade Learning Environment: An Evaluation Platform for General Agents | ✓ Link | 628 | Best Learner | 2012-07-19 |
| CURL: Contrastive Unsupervised Representations for Reinforcement Learning | ✓ Link | 543.7 | CURL | 2020-04-08 |
| []() | | 537.0 | SARSA | |
| Soft Actor-Critic for Discrete Action Settings | ✓ Link | 350 | SAC | 2019-10-16 |