| Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model | ✓ Link | 260.13 | MuZero | 2019-11-19 |
| First return, then explore | ✓ Link | 260 | Go-Explore | 2020-04-27 |
| Agent57: Outperforming the Atari Human Benchmark | ✓ Link | 251.18 | Agent57 | 2020-03-30 |
| Recurrent Experience Replay in Distributed Reinforcement Learning | ✓ Link | 219.5 | R2D2 | 2019-05-01 |
| Generalized Data Distribution Iteration | | 205.2 | GDI-H3 | 2022-06-07 |
| GDI: Rethinking What Makes Reinforcement Learning Different From Supervised Learning | | 201.9 | GDI-I3 | 2021-06-11 |
| Generalized Data Distribution Iteration | | 201.9 | GDI-I3 | 2022-06-07 |
| DNA: Proximal Policy Optimization with a Dual Network Architecture | ✓ Link | 181 | DNA | 2022-06-20 |
| RUDDER: Return Decomposition for Delayed Rewards | ✓ Link | 179 | RUDDER | 2018-06-20 |
| Online and Offline Reinforcement Learning by Planning with a Learned Model | ✓ Link | 131.65 | MuZero (Res2 Adam) | 2021-04-13 |
| Fully Parameterized Quantile Function for Distributional Reinforcement Learning | ✓ Link | 102.3 | FQF | 2019-11-05 |
| Learning values across many orders of magnitude | | 102.1 | DDQN+Pop-Art noop | 2016-02-24 |
| Implicit Quantile Networks for Distributional Reinforcement Learning | ✓ Link | 86.5 | IQN | 2018-06-14 |
| Evolving simple programs for playing Atari games | ✓ Link | 85.8 | CGP | 2018-06-14 |
| A Distributional Perspective on Reinforcement Learning | ✓ Link | 81.8 | C51 noop | 2017-07-21 |
| The Reactor: A fast and sample-efficient Actor-Critic agent for Reinforcement Learning | | 81.0 | Reactor 500M | 2017-04-15 |
| Distributional Reinforcement Learning with Quantile Regression | ✓ Link | 77.2 | QR-DQN-1 | 2017-10-27 |
| Increasing the Action Gap: New Operators for Reinforcement Learning | ✓ Link | 71.59 | Persistent AL | 2015-12-15 |
| Deep Reinforcement Learning with Double Q-learning | ✓ Link | 69.6 | DDQN (tuned) hs | 2015-09-22 |
| Dueling Network Architectures for Deep Reinforcement Learning | ✓ Link | 68.1 | DDQN (tuned) noop | 2015-11-20 |
| Dueling Network Architectures for Deep Reinforcement Learning | ✓ Link | 65.7 | Duel hs | 2015-11-20 |
| Dueling Network Architectures for Deep Reinforcement Learning | ✓ Link | 65.5 | Duel noop | 2015-11-20 |
| Train a Real-world Local Path Planner in One Hour via Partially Decoupled Reinforcement Learning and Vectorized Diversity | ✓ Link | 62.4 | ASL DDQN | 2023-05-07 |
| Deep Exploration via Bootstrapped DQN | ✓ Link | 60.2 | Bootstrapped DQN | 2016-02-15 |
| IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures | ✓ Link | 59.92 | IMPALA (deep) | 2018-02-05 |
| Increasing the Action Gap: New Operators for Reinforcement Learning | ✓ Link | 57.41 | Advantage Learning | 2015-12-15 |
| Deep Reinforcement Learning with Double Q-learning | ✓ Link | 56.5 | DQN hs | 2015-09-22 |
| Massively Parallel Methods for Deep Reinforcement Learning | ✓ Link | 54 | Gorila | 2015-07-15 |
| Prioritized Experience Replay | ✓ Link | 52 | Prior hs | 2015-11-18 |
| Deep Reinforcement Learning with Double Q-learning | ✓ Link | 50.4 | DQN noop | 2015-09-22 |
| Deep Reinforcement Learning with Double Q-learning | ✓ Link | 50.4 | Prior+Duel hs | 2015-09-22 |
| Mastering Atari with Discrete World Models | ✓ Link | 49 | DreamerV2 | 2020-10-05 |
| Prioritized Experience Replay | ✓ Link | 47.9 | Prior noop | 2015-11-18 |
| Dueling Network Architectures for Deep Reinforcement Learning | ✓ Link | 46.7 | Prior+Duel noop | 2015-11-20 |
| The Arcade Learning Environment: An Evaluation Platform for General Agents | ✓ Link | 43.9 | Best Learner | 2012-07-19 |
| Human level control through deep reinforcement learning | ✓ Link | 42.4 | Nature DQN | 2015-02-25 |
| Asynchronous Methods for Deep Reinforcement Learning | ✓ Link | 41.8 | A3C LSTM hs | 2016-02-04 |
| Policy Optimization With Penalized Point Probability Distance: An Alternative To Proximal Policy Optimization | ✓ Link | 38.99 | POP3D | 2018-07-02 |
| | 36.4 | SARSA | |
| Asynchronous Methods for Deep Reinforcement Learning | ✓ Link | 36.2 | A3C FF (1 day) hs | 2016-02-04 |
| Asynchronous Methods for Deep Reinforcement Learning | ✓ Link | 35.1 | A3C FF hs | 2016-02-04 |
| Self-Imitation Learning | ✓ Link | 31.1 | A2C + SIL | 2018-06-14 |
| Evolution Strategies as a Scalable Alternative to Reinforcement Learning | ✓ Link | 30 | ES FF (1 hour) noop | 2017-03-10 |
| Distributed Prioritized Experience Replay | ✓ Link | 17.6 | Ape-X | 2018-03-02 |