Sparse Reward Design

Sparse Reward Design is a reward engineering methodology in reinforcement learning wherein the reward signal is deliberately structured to be provided only at sparse intervals or upon achievement of specific milestones, distinguishing it from dense reward designs by its minimal and infrequent feedback to guide agent learning.

43

Publications

1.4K

Citations

201

Authors

74

Institutions

Publications per year

2017–2026

42

Authors

201

  1. YW
    Yuechen Wu

    University of Arizona, United States

    2 publications · 66 citations

    2

    h-index

  2. WZ
    Wei Zhang

    Shandong University, China

    2 publications · 66 citations

    2

    h-index

  3. RW
    Rui Wang

    National University of Defense Technology, China

    2 publications · 64 citations

    2

    h-index

  4. CZ
    Chongjie Zhang

    Tsinghua University, China

    2 publications · 64 citations

    2

    h-index

  5. SL
    Shijian Lu

    Nanyang Technological University, Singapore

    2 publications · 66 citations

    2

    h-index

1–5 of 201

Venues

  1. arXiv (Cornell University)

    Repository · 9 publications · 285 citations

    9

    h-index

  2. 3

    h-index

  3. Proceedings of the AAAI Conference on Artificial Intelligence

    Conference · 3 publications · 74 citations

    3

    h-index

  4. Machines

    2 publications · 62 citations

    2

    h-index

  5. IEEE Internet of Things Journal

    2 publications · 93 citations

    2

    h-index

1–5