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

    Neuronpedia

    INDEX
    Explanations

    code

    np_max-act · gemini-2.0-flash

    The neuron detects template placeholder markers—specifically the “[Here …]” segments that introduce where to fill in recipe name, ingredients, or steps.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     stejně
    -0.07
    _money
    -0.07
     BoxFit
    -0.07
     Baghd
    -0.07
    .ReadString
    -0.07
    /utility
    -0.07
     wenig
    -0.06
     работе
    -0.06
     desperately
    -0.06
    &display
    -0.06
    POSITIVE LOGITS
    duct
    0.07
     photo
    0.07
    Vol
    0.07
    ॉर
    0.06
     multiple
    0.06
     Multiple
    0.06
    Tax
    0.06
     Dir
    0.06
    INNER
    0.06
    Creative
    0.06
    Act Density 0.004%

    No Known Activations