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

    Neuronpedia

    INDEX
    Explanations

    expressing information

    np_max-act · gemini-2.0-flash

    This neuron detects action or command verbs in user instructions (e.g. “access,” “say,” “generate,” “display”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    .INVISIBLE
    -0.07
    ORA
    -0.06
    (Art
    -0.06
    sendKeys
    -0.06
     Focus
    -0.06
     か
    -0.06
     Eins
    -0.06
     svůj
    -0.06
    ")[
    -0.06
    .Keyboard
    -0.06
    POSITIVE LOGITS
     Classics
    0.08
     guts
    0.07
    -derived
    0.07
    ousse
    0.07
     sorts
    0.06
     simplified
    0.06
    füg
    0.06
    :?
    0.06
     Suggestions
    0.06
    ρίς
    0.06
    Act Density 0.003%

    No Known Activations