the open
    research
    Contribute
    Back to the reading lists

    Safety, alignment & interpretability· 2020

    Concept Bottleneck Models

    Pang Wei Koh (first author) · 2020

    Read the original

    arXiv 2007.04612OpenAlex W3035602609

    On the reading lists

    Why it matters

    Made predictions depend explicitly on human-interpretable intermediate concepts, enabling targeted intervention.

    Condensed from the expert lists' notes.

    The record

    First released
    2020 (from arXiv 2007.04612)
    Citations
    1,559 citations, as counted by Papers With Code (CC BY-SA 4.0)
    Named by
    both expert lists
    Score
    0.341 · computed rank 503 · consensus 0.49, must-read 0, citations percentile 0.553 (2020s · safety), awards 0, existing lists 0. Method

    Data sources

    Catalogue data from OpenAlex (CC0), Crossref (CC0 metadata; no abstracts are shown), arXiv, and Papers With Code (CC BY-SA 4.0). Each count names its source. See the method.