the open
    research
    Contribute
    Back to the reading lists

    Safety, alignment & interpretability· 2019

    Are Sixteen Heads Really Better than One?

    Paul Michel (first author) · 2019

    Read the original

    arXiv 1905.10650OpenAlex W2945767825

    On the reading lists

    Why it matters

    Showed many attention heads are redundant.

    Condensed from the expert lists' notes.

    The record

    First released
    2019 (from arXiv 1905.10650)
    Citations
    1,550 citations, as counted by Papers With Code (CC BY-SA 4.0)
    Named by
    the grok list only
    Score
    0.274 · computed rank 669 · consensus 0.23, must-read 0, citations percentile 0.585 (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.