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

    Neuronpedia

    INDEX
    Explanations

    percentages or portions of a whole as represented by the verb "accounted for."

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to accounting and percentages

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    fml
    -0.70
    iev
    -0.69
    ASY
    -0.69
    iw
    -0.67
    Lab
    -0.67
    rett
    -0.65
    Attempts
    -0.65
    ieties
    -0.64
    RIP
    -0.63
     awaits
    -0.62
    POSITIVE LOGITS
    illac
    0.85
     some
    0.80
     alot
    0.79
     most
    0.76
     several
    0.76
     nearly
    0.76
     billions
    0.74
     many
    0.73
     both
    0.72
    anni
    0.72
    Act Density 0.070%

    No Known Activations