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

    Neuronpedia

    INDEX
    Explanations

    phrases related to rules and regulations being applied or not applied in different contexts

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "apply" and its variations in legal or regulatory contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    asted
    -0.77
    airs
    -0.74
    ument
    -0.73
    ifa
    -0.71
    selage
    -0.70
    ighters
    -0.70
    obs
    -0.70
    buck
    -0.70
    roman
    -0.70
    orr
    -0.68
    POSITIVE LOGITS
     apply
    0.85
     applies
    0.81
    rences
    0.78
     uniformly
    0.77
     universally
    0.76
     applied
    0.76
     inconsist
    0.75
    ãģĨ
    0.75
    alties
    0.73
     faithfully
    0.72
    Act Density 0.012%

    No Known Activations