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

    Neuronpedia

    INDEX
    Explanations

    pronouns in sentences where the subject is being observed or addressed

    oai_token-act-pair · gpt-3.5-turbo

    references to personal relationships and interactions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ģĸ
    -0.68
    icion
    -0.65
    ãĤ¦ãĤ¹
    -0.63
    creation
    -0.63
    é¾
    -0.62
    oso
    -0.62
    elsen
    -0.62
    Questions
    -0.60
    osate
    -0.60
    osi
    -0.59
    POSITIVE LOGITS
     smiling
    1.00
     naked
    0.99
     alive
    0.95
     interact
    0.93
     perform
    0.91
     behaving
    0.89
     dancing
    0.88
     bleed
    0.86
     unfold
    0.86
     interacting
    0.84
    Act Density 0.151%

    No Known Activations