the open
    research
    Contribute
    Back to the reading lists

    Safety, alignment & interpretability· 2017

    Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)

    Been Kim (first author) · 2017 · published 2018

    Read the original

    arXiv 1711.11279OpenAlex W2796885425

    On the reading lists

    Why it matters

    Shifted explanations from individual features toward human-understandable concepts and quantitative sensitivity tests.

    Condensed from the expert lists' notes.

    The record

    First released
    2017 (from arXiv 1711.11279)
    Published
    2018
    Citations
    2,563 citations, as counted by Papers With Code (CC BY-SA 4.0)
    Named by
    both expert lists
    Score
    0.544 · computed rank 197 · consensus 0.89, must-read 0.5, citations percentile 0.649 (2010s · 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.