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

    Neuronpedia

    INDEX
    Explanations

    statements of regret or hypothetical scenarios in past tense

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Negative Logits
     supposedly
    -0.80
     purportedly
    -0.74
     allegedly
    -0.70
     currently
    -0.70
    osi
    -0.64
    uras
    -0.61
     encl
    -0.58
     promise
    -0.58
     portray
    -0.58
     diagnose
    -0.58
    POSITIVE LOGITS
     sooner
    0.80
    fall
    0.79
    fitting
    0.76
    ivably
    0.69
     forgiven
    0.68
    acted
    0.68
     tomorrow
    0.67
     someday
    0.67
    argon
    0.62
    header
    0.62
    Act Density 3.939%

    No Known Activations