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

    Neuronpedia

    INDEX
    Explanations

    phrases related to legal or investigative contexts, possibly focusing on actions, knowledge, and intentions

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to legal actions and justifications

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     partName
    -0.80
    iku
    -0.72
    icles
    -0.68
    icle
    -0.68
    lator
    -0.68
    ftime
    -0.67
    ikan
    -0.65
    phies
    -0.65
    Zone
    -0.64
    otiation
    -0.63
    POSITIVE LOGITS
     unlawfully
    1.11
     improper
    1.04
     improperly
    1.03
     unlawful
    1.01
     illegally
    0.98
     lawfully
    0.97
     wiret
    0.93
     inappropriately
    0.89
     wrongdoing
    0.88
     violated
    0.86
    Act Density 0.521%

    No Known Activations