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

    Neuronpedia

    INDEX
    Explanations

    the word "with" followed by the potential object of the interaction

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the phrase "speak with."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    fighter
    -0.75
    ynski
    -0.66
    mol
    -0.64
    atra
    -0.63
    bourg
    -0.62
    waters
    -0.61
    coming
    -0.60
    gross
    -0.60
    DIV
    -0.59
    hra
    -0.58
    POSITIVE LOGITS
     regards
    0.96
     regard
    0.90
     representatives
    0.88
     assistants
    0.77
     executives
    0.77
     constituents
    0.77
     coworkers
    0.76
     stakeholders
    0.75
     him
    0.74
     fellow
    0.73
    Act Density 0.099%

    No Known Activations