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

    Neuronpedia

    INDEX
    Explanations

    phrases indicating a change or comparison from how things used to be

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating past states or conditions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     defy
    -0.78
     seize
    -0.74
    âĦ¢:
    -0.74
    udge
    -0.72
     traverse
    -0.70
     compose
    -0.70
     cease
    -0.69
     claim
    -0.69
     attempt
    -0.68
     result
    -0.68
    POSITIVE LOGITS
    hemoth
    0.96
     able
    0.96
    leeve
    0.89
    fits
    0.86
    league
    0.82
     regarded
    0.81
    held
    0.79
     judged
    0.78
     considered
    0.75
     treated
    0.74
    Act Density 0.078%

    No Known Activations