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

    Neuronpedia

    INDEX
    Explanations

    the word "which" and its variations, indicating questions or inquiries about specific choices or options

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    Ú¯ÛĮر
    -0.15
    ust
    -0.15
    .codehaus
    -0.14
    loff
    -0.14
    egg
    -0.14
    ÑģÑĤÑĭ
    -0.14
    sson
    -0.14
    ÏĥÏĦα
    -0.13
    ald
    -0.13
    ãģĦãģĨ
    -0.13
    POSITIVE LOGITS
    soever
    0.31
     ones
    0.30
     direction
    0.27
     way
    0.25
    -ever
    0.24
     among
    0.24
    -way
    0.23
    /how
    0.23
    ones
    0.21
    among
    0.21
    Act Density 0.046%

    No Known Activations