Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    the word "which," indicating it is focused on clauses or phrases that provide additional information or clarification

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ey
    -0.16
    ære
    -0.15
    igraph
    -0.15
    wor
    -0.15
     what
    -0.15
    ouv
    -0.14
    eye
    -0.14
    ish
    -0.14
    ivor
    -0.14
    ikel
    -0.13
    POSITIVE LOGITS
    soever
    0.34
     we
    0.17
    itzer
    0.17
    oping
    0.17
    chaft
    0.16
    ÑģÑĮ
    0.16
    pring
    0.15
    ady
    0.15
    andler
    0.15
     they
    0.14
    Act Density 0.046%

    No Known Activations