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

    Neuronpedia

    INDEX
    Explanations

    personal interactions involving giving or receiving advice or directives

    oai_token-act-pair · gpt-3.5-turbo

    pronouns used in directives or communication

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    arious
    -0.61
    aired
    -0.61
    arij
    -0.60
    hement
    -0.60
    arette
    -0.59
    hift
    -0.59
    ibal
    -0.59
    pei
    -0.59
    atl
    -0.59
    xtap
    -0.58
    POSITIVE LOGITS
     beforehand
    0.89
     goodbye
    0.83
     how
    0.82
     bluntly
    0.80
     orally
    0.79
     plainly
    0.77
     about
    0.77
     why
    0.77
     otherwise
    0.73
    ledge
    0.70
    Act Density 0.079%

    No Known Activations