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

    Neuronpedia

    INDEX
    Explanations

    comparisons or evaluations followed by a positive or negative sentiment

    oai_token-act-pair · gpt-3.5-turbo

    phrases that assert a state or identity

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    itches
    -0.70
    ffe
    -0.70
    actory
    -0.65
    atre
    -0.64
    ees
    -0.62
    ievers
    -0.61
    gg
    -0.61
    lex
    -0.60
     Gir
    -0.60
    agon
    -0.60
    POSITIVE LOGITS
     admittedly
    1.02
     comprised
    0.94
     basically
    0.92
     supposed
    0.91
     essentially
    0.90
     meant
    0.89
     unlikely
    0.88
     undoubtedly
    0.87
     presumably
    0.86
     obviously
    0.86
    Act Density 0.164%

    No Known Activations