| First return, then explore | ✓ Link | 95756 | Go-Explore | 2020-04-27 |
| Agent57: Outperforming the Atari Human Benchmark | ✓ Link | 79716.46 | Agent57 | 2020-03-30 |
| Self-supervised network distillation: an effective approach to exploration in sparse reward environments | ✓ Link | 17313 | SND-VIC | 2023-02-22 |
| Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model | ✓ Link | 15299.98 | MuZero | 2019-11-19 |
| GDI: Rethinking What Makes Reinforcement Learning Different From Supervised Learning | | 15100 | GDI-I3 | 2021-06-11 |
| Generalized Data Distribution Iteration | | 15100 | GDI-I3 | 2022-06-07 |
| Generalized Data Distribution Iteration | | 15100 | GDI-H3 | 2022-06-07 |
| A Distributional Perspective on Reinforcement Learning | ✓ Link | 15095.0 | C51 noop | 2017-07-21 |
| Self-supervised network distillation: an effective approach to exploration in sparse reward environments | ✓ Link | 15089 | SND-STD | 2023-02-22 |
| Evolving simple programs for playing Atari games | ✓ Link | 12702.2 | CGP | 2018-06-14 |
| Exploration by Random Network Distillation | ✓ Link | 8666 | RND | 2018-10-30 |
| Count-Based Exploration with Neural Density Models | ✓ Link | 8358.7 | DQN-PixelCNN | 2017-03-03 |
| Recurrent Experience Replay in Distributed Reinforcement Learning | ✓ Link | 5322.7 | R2D2 | 2019-05-01 |
| Increasing the Action Gap: New Operators for Reinforcement Learning | ✓ Link | 5276.16 | Advantage Learning | 2015-12-15 |
| Self-supervised network distillation: an effective approach to exploration in sparse reward environments | ✓ Link | 4213 | SND-V | 2023-02-22 |
| Large-Scale Study of Curiosity-Driven Learning | ✓ Link | 3036.5 | Intrinsic Reward Agent | 2018-08-13 |
| Massively Parallel Methods for Deep Reinforcement Learning | ✓ Link | 2598.6 | Gorila | 2015-07-15 |
| Mastering Atari with Discrete World Models | ✓ Link | 2198 | DreamerV2 | 2020-10-05 |
| The Arcade Learning Environment: An Evaluation Platform for General Agents | ✓ Link | 1947.3 | Best Baseline | 2012-07-19 |
| Deep Exploration via Bootstrapped DQN | ✓ Link | 1812.5 | Bootstrapped DQN | 2016-02-15 |
| Human level control through deep reinforcement learning | ✓ Link | 1788.0 | Nature DQN | 2015-02-25 |
| Deep Reinforcement Learning with Double Q-learning | ✓ Link | 1277.6 | Prior+Duel hs | 2015-09-22 |
| The Arcade Learning Environment: An Evaluation Platform for General Agents | ✓ Link | 684.3 | Best Learner | 2012-07-19 |
| Prioritized Experience Replay | ✓ Link | 670.7 | Prior hs | 2015-11-18 |
| Self-Imitation Learning | ✓ Link | 661.2 | A2C + SIL | 2018-06-14 |
| Asynchronous Methods for Deep Reinforcement Learning | ✓ Link | 421.1 | A3C LSTM hs | 2016-02-04 |
| Distributional Reinforcement Learning with Quantile Regression | ✓ Link | 350 | QR-DQN-1 | 2017-10-27 |
| Train a Real-world Local Path Planner in One Hour via Partially Decoupled Reinforcement Learning and Vectorized Diversity | ✓ Link | 349.7 | ASL DDQN | 2023-05-07 |
| Dueling Network Architectures for Deep Reinforcement Learning | ✓ Link | 292.6 | Duel hs | 2015-11-20 |
| Learning values across many orders of magnitude | | 286.7 | DDQN+Pop-Art noop | 2016-02-24 |
| Noisy Networks for Exploration | ✓ Link | 279 | NoisyNet-Dueling | 2017-06-30 |
| Deep Reinforcement Learning with Double Q-learning | ✓ Link | 207.9 | DQN hs | 2015-09-22 |
| Asynchronous Methods for Deep Reinforcement Learning | ✓ Link | 206.9 | A3C FF hs | 2016-02-04 |
| Dueling Network Architectures for Deep Reinforcement Learning | ✓ Link | 206.0 | Prior+Duel noop | 2015-11-20 |
| Count-Based Exploration with Neural Density Models | ✓ Link | 206.0 | DQN-CTS | 2017-03-03 |
| Prioritized Experience Replay | ✓ Link | 200.0 | Prior noop | 2015-11-18 |
| Implicit Quantile Networks for Distributional Reinforcement Learning | ✓ Link | 200 | IQN | 2018-06-14 |
| Asynchronous Methods for Deep Reinforcement Learning | ✓ Link | 194.4 | A3C FF (1 day) hs | 2016-02-04 |
| Deep Reinforcement Learning with Double Q-learning | ✓ Link | 146.7 | DQN noop | 2015-09-22 |
| Dueling Network Architectures for Deep Reinforcement Learning | ✓ Link | 129.7 | DDQN (tuned) noop | 2015-11-20 |
| CURL: Contrastive Unsupervised Representations for Reinforcement Learning | ✓ Link | 105.2 | CURL | 2020-04-08 |
| Dueling Network Architectures for Deep Reinforcement Learning | ✓ Link | 103.0 | Duel noop | 2015-11-20 |
| Evolution Strategies as a Scalable Alternative to Reinforcement Learning | ✓ Link | 100.0 | ES FF (1 hour) noop | 2017-03-10 |
| Online and Offline Reinforcement Learning by Planning with a Learned Model | ✓ Link | 100 | MuZero (Res2 Adam) | 2021-04-13 |
| DNA: Proximal Policy Optimization with a Dual Network Architecture | ✓ Link | 100 | DNA | 2022-06-20 |
| Unifying Count-Based Exploration and Intrinsic Motivation | ✓ Link | 99.32 | A3C-CTS | 2016-06-06 |
| Count-Based Exploration with the Successor Representation | ✓ Link | 99.1 | DQNMMCe+SR | 2018-07-31 |
| IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures | ✓ Link | 98.50 | IMPALA (deep) | 2018-02-05 |
| []() | | 86.0 | SARSA | |
| Policy Optimization With Penalized Point Probability Distance: An Alternative To Proximal Policy Optimization | ✓ Link | 79.67 | POP3D | 2018-07-02 |
| Distributed Prioritized Experience Replay | ✓ Link | 49.8 | Ape-X | 2018-03-02 |
| Deep Reinforcement Learning with Double Q-learning | ✓ Link | -575.5 | DDQN (tuned) hs | 2015-09-22 |