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

    Neuronpedia

    INDEX
    Explanations

    change

    np_max-act · gemini-2.0-flash

    This neuron detects discourse markers that signal a change or reversal in state (e.g. “no longer,” “has changed,” “now”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    Why
    -0.07
    eken
    -0.07
     heel
    -0.07
     radius
    -0.06
     grep
    -0.06
    Equivalent
    -0.06
     получения
    -0.06
    ih
    -0.06
    Retry
    -0.06
     recipro
    -0.06
    POSITIVE LOGITS
     dece
    0.06
     zajist
    0.06
    asmine
    0.06
     Львів
    0.06
     footsteps
    0.06
     pří
    0.06
     Odkazy
    0.06
     шлях
    0.06
    _fonts
    0.06
    ंप
    0.06
    Act Density 0.034%

    No Known Activations