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

    Neuronpedia

    INDEX
    Explanations

    interrogative words or phrases that suggest questions or conditions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    question and command starters

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron detects sentence-initial interrogative or conditional clause starters—i.e., the beginnings of questions or conditionals.

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Negative Logits
    parsedMessage
    -0.88
    adaptiveStyles
    -0.82
    setVerticalGroup
    -0.76
    windowFixed
    -0.72
    oredCriteria
    -0.71
     незавершена
    -0.66
    addCriterion
    -0.65
    SourceChecksum
    -0.65
    msgTypes
    -0.65
    Geplaatst
    -0.64
    POSITIVE LOGITS
     inalámbrico
    0.40
    mathrm
    0.38
     literario
    0.36
     deportivo
    0.35
     Edad
    0.34
     vejo
    0.33
     tuya
    0.33
    atória
    0.33
     暮らし
    0.33
     división
    0.32
    Act Density 0.011%

    No Known Activations