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

    Neuronpedia

    INDEX
    Explanations

    words related to decision making or actions that need to be taken

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "to" and similar infinitive verbs or constructs indicating future action or necessity

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     hates
    -0.72
     cursing
    -0.71
     slaying
    -0.70
     tossing
    -0.70
     cultivating
    -0.70
     imagining
    -0.69
     searching
    -0.69
     telling
    -0.69
     hoping
    -0.68
     dreaming
    -0.68
    POSITIVE LOGITS
     occur
    1.60
     happen
    1.39
     be
    1.38
     arise
    1.26
     exist
    1.22
     remain
    1.21
     expire
    1.18
     emerge
    1.17
     proceed
    1.15
     originate
    1.12
    Act Density 0.286%

    No Known Activations