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

    Neuronpedia

    INDEX
    Explanations

    references to being taken or held under some form of custody or restriction

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "detained."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    hl
    -0.76
    clerosis
    -0.72
    vironment
    -0.70
    soDeliveryDate
    -0.70
    jri
    -0.70
    ocaly
    -0.67
    Guide
    -0.65
    bold
    -0.61
    iosyncr
    -0.61
    shadow
    -0.60
    POSITIVE LOGITS
     detained
    1.02
     detain
    0.98
     detainees
    0.89
     prisoners
    0.78
     detention
    0.74
     Cust
    0.72
     hostages
    0.71
     handcuffed
    0.70
     aback
    0.69
    uan
    0.68
    Act Density 0.018%

    No Known Activations