Meta-reinforcement Learning

Meta-Reinforcement Learning is a meta-learning framework that enables an agent to acquire generalized reinforcement learning strategies across diverse tasks by learning how to adapt its policy or value functions through experience over multiple environments or episodes.

192

Publications

10.1K

Citations

727

Authors

231

Institutions

Publications per year

2017–2026

186

Authors

727

  1. SL
    Sergey Levine

    University of California, Berkeley, United States

    19 publications · 1.8K citations

    17

    h-index

  2. CF
    Chelsea Finn

    University of California, Berkeley, United States

    17 publications · 1.5K citations

    16

    h-index

  3. PA
    Pieter Abbeel

    University of California, Berkeley, United States

    12 publications · 959 citations

    11

    h-index

  4. HV
    Hado van Hasselt

    Google DeepMind (United Kingdom), United Kingdom

    6 publications · 347 citations

    6

    h-index

  5. JL
    Jiawen Li

    South China University of Technology, China

    6 publications · 190 citations

    6

    h-index

1–5 of 727

Venues

  1. arXiv (Cornell University)

    Repository · 56 publications · 3.3K citations

    26

    h-index

  2. IEEE Internet of Things Journal

    7 publications · 315 citations

    7

    h-index

  3. Acta Astronautica

    5 publications · 306 citations

    5

    h-index

  4. Neural Information Processing Systems

    Conference · 5 publications · 121 citations

    5

    h-index

  5. IEEE Access

    4 publications · 81 citations

    4

    h-index

1–5