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

    Neuronpedia

    INDEX
    Explanations

    modal verbs and expressions of potentiality or obligation

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    uje
    -0.15
    ouch
    -0.14
     underwent
    -0.14
    ãģĭãĤīãģ¯
    -0.14
    asl
    -0.14
    113
    -0.13
    took
    -0.13
    éĥ½ä¼ļ
    -0.13
     undergo
    -0.13
    lds
    -0.12
    POSITIVE LOGITS
     be
    0.56
     бÑĭÑĤÑĮ
    0.46
     být
    0.42
     being
    0.42
     been
    0.39
     бÑĥÑĤи
    0.38
     essere
    0.37
     byÄĩ
    0.37
    being
    0.36
     être
    0.34
    Act Density 0.272%

    No Known Activations