{"id":"arxiv:1712.06567","title":"Deep Neuroevolution: Genetic Algorithms Are a Competitive Alternative for Training Deep Neural Networks for Reinforcement Learning","date":"2017-12-18","section":"Reinforcement learning","quality":0.395745,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/","source_url":"https://arxiv.org/abs/1712.06567","text_available":true,"authors":["Felipe Petroski Such","Vashisht Madhavan","Edoardo Conti","Joel Lehman"],"outgoing":[{"source":"arxiv:1712.06567","target":"arxiv:1509.06461","evidence":"title","id":"arxiv:1509.06461","title":"Deep Reinforcement Learning with Double Q-learning","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/a2f115ac879cd2efb9548baf3ecdcb925d2492f03b4b0940720f6e5b0840715b/"},{"source":"arxiv:1712.06567","target":"arxiv:1511.05952","evidence":"id","id":"arxiv:1511.05952","title":"Prioritized Experience Replay","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/c88bf0c11c1817dcc457c001dfef5cd120a0946c8a009a0fe40648dae03eb7d4/"},{"source":"arxiv:1712.06567","target":"arxiv:1511.06581","evidence":"id","id":"arxiv:1511.06581","title":"Dueling Network Architectures for Deep Reinforcement Learning","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/0a5f0593cf943c782ba38ae2fc386f953f859b9d3fc9722e8f17f67ca9cf50b5/"},{"source":"arxiv:1712.06567","target":"arxiv:1707.06887","evidence":"id","id":"arxiv:1707.06887","title":"A Distributional Perspective on Reinforcement Learning","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/96b040bb771d28ca3254b5519a2191c3e958a35dfdd60075344ac86af6a590e1/"},{"source":"arxiv:1712.06567","target":"arxiv:1710.02298","evidence":"id","id":"arxiv:1710.02298","title":"Rainbow: Combining Improvements in Deep Reinforcement Learning","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/4183a5aecdb6558733bd0b2028e6197904c8b46173181bf6c8270fb6cbcfc973/"}],"incoming":[{"source":"arxiv:1905.10985","target":"arxiv:1712.06567","evidence":"id","id":"arxiv:1905.10985","title":"AI-GAs: AI-generating algorithms, an alternate paradigm for producing general artificial intelligence","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/089142f7297bed414d01b82c78dc888e4018fecff7efbe5cae66442e03f87c4b/"},{"source":"arxiv:2004.12919","target":"arxiv:1712.06567","evidence":"id","id":"arxiv:2004.12919","title":"First return, then explore","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/3cd790c501eaa633193918533947c58e30d1939bc08577606f1d88c56d368285/"},{"source":"arxiv:2601.20861","target":"arxiv:1712.06567","evidence":"id","id":"arxiv:2601.20861","title":"Evolutionary Strategies lead to Catastrophic Forgetting in LLMs","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/4178758b0cb3ebc95f6a3cbc6423e8ea05c0ae649e11a2f4059a84b3faa6cde0/"},{"source":"doi:10.1038/s41467-021-26568-2","target":"arxiv:1712.06567","evidence":"id","id":"doi:10.1038/s41467-021-26568-2","title":"Correspondence between neuroevolution and gradient descent","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/21875123ceb4116f9890b5b8d70154234ec0aa1bb29c70d3439a48aad10629b0/"}],"text_parts":[{"start":0,"end":15921,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-1.html"},{"start":15921,"end":31297,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-2.html"},{"start":31297,"end":46638,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-3.html"},{"start":46638,"end":51895,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-4.html"}],"contents":[{"title":"Deep Neuroevolution: Genetic Algorithms are a Competitive Alternative for Training Deep Neural Networks for Reinforcement Learning","offset":0,"level":1,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-1.html#text"},{"title":"Abstract","offset":134,"level":6,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-1.html#text"},{"title":"1 Introduction","offset":1982,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-1.html#text"},{"title":"2 Background","offset":6424,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-1.html#text"},{"title":"3 Methods","offset":9240,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-1.html#text"},{"title":"3.1 Genetic Algorithm","offset":9254,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-1.html#text"},{"title":"3.2 Novelty Search","offset":14948,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-1.html#text"},{"title":"4 Experiments","offset":16858,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-2.html#text"},{"title":"4.1 Atari","offset":18078,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-2.html#text"},{"title":"4.2 Image Hard Maze","offset":28334,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-2.html#text"},{"title":"4.3 Humanoid Locomotion","offset":33610,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-3.html#text"},{"title":"5 Discussion","offset":34173,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-3.html#text"},{"title":"6 Conclusion","offset":39287,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-3.html#text"},{"title":"Acknowledgements","offset":40524,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-3.html#text"},{"title":"7 Supplementary Information","offset":40968,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-3.html#text"},{"title":"7.1 Why the GA is faster than ES","offset":41000,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-3.html#text"},{"title":"7.2 Hyperparameters","offset":42211,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-3.html#text"},{"title":"7.3 Additional information about the Deep GA compact encoding method","offset":42555,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-3.html#text"},{"title":"7.4 Additional experimental details for the Image Hard Maze","offset":44307,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-3.html#text"},{"title":"7.5 Humanoid Locomotion","offset":45223,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-3.html#text"},{"title":"7.6 The meaning of “frames”","offset":49410,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-4.html#text"},{"title":"Instructions for reporting errors","offset":50931,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/part-4.html#text"}],"text_code_points":51895,"source_markdown":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7191dc11e224beb3dc9abb8cd4efc58c70b79939b61fd56e4416b4b5e1404312/source.md"}