Concept-based Explanations

Concept-Based Explanations is a type of interpretable machine learning method that identifies high-level human-understandable concepts—such as object parts or semantic attributes—and quantifies their contributions to a model's predictions, thereby explaining model behavior in terms of meaningful data abstractions rather than raw features.

110

Publications

43.8K

Citations

344

Authors

144

Institutions

Publications per year

2017–2026

78

Authors

344

  1. BK
    Been Kim

    Google (United States), United States

    6 publications · 315 citations

    6

    h-index

  2. PL
    Píetro Lió

    University of Cambridge, United Kingdom

    5 publications · 160 citations

    5

    h-index

  3. SL
    Sebastian Lapuschkin

    Fraunhofer Institute for Telecommunications, Heinrich Hertz Institute, Germany

    3 publications · 189 citations

    3

    h-index

  4. PB
    Pietro Barbiero

    University of Cambridge, United Kingdom

    3 publications · 123 citations

    3

    h-index

  5. TL
    Tania Lombrozo

    University of California, Berkeley, United States

    3 publications · 131 citations

    3

    h-index

1–5 of 344

Venues

  1. arXiv (Cornell University)

    Repository · 17 publications · 633 citations

    13

    h-index

  2. Proceedings of the AAAI Conference on Artificial Intelligence

    Conference · 6 publications · 273 citations

    6

    h-index

  3. IEEE Access

    3 publications · 177 citations

    3

    h-index

  4. Cognition

    3 publications · 152 citations

    3

    h-index

  5. 2

    h-index

1–5