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

    Neuronpedia

    INDEX
    Explanations

    politely stated phrases or requests

    oai_token-act-pair · gpt-3.5-turbo

    expressions of apology and request for politeness

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     Liberty
    -0.82
     Caribbean
    -0.75
     Rockefeller
    -0.73
     NYC
    -0.71
     pdf
    -0.70
     telecom
    -0.69
     retirees
    -0.68
     Berm
    -0.68
     centrally
    -0.68
     Tourism
    -0.67
    POSITIVE LOGITS
    âĢ
    1.73
    ··
    1.48
    âĻ
    1.27
    ?ãĢį
    1.27
    âĶĢ
    1.22
    âĶĢâĶĢ
    1.22
    âĢķ
    1.22
    .ãĢį
    1.15
    âĶĢâĶĢâĶĢâĶĢ
    1.11
    .</
    1.10
    Act Density 0.640%

    No Known Activations