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

    Neuronpedia

    INDEX
    Explanations

    About

    np_max-act · gemini-2.0-flash

    The neuron consistently lights up on tokens that are part of the “Document:” block itself—that is, it marks the actual source‐text lines after the “Document:” header.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    dialogue involving requests for assistance and responses related to performance or well-being.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    Yellow
    -0.08
    Warnings
    -0.07
    кет
    -0.07
    ůž
    -0.06
     Yellow
    -0.06
    ulpt
    -0.06
     opioid
    -0.06
     Seahawks
    -0.06
     yellow
    -0.06
     Gupta
    -0.06
    POSITIVE LOGITS
    :::::::::::
    0.07
     ERA
    0.07
    /arm
    0.07
    .Empty
    0.07
     sta
    0.07
     io
    0.06
    ]int
    0.06
    limits
    0.06
     。
    0.06
    .setInt
    0.06
    Act Density 0.014%

    No Known Activations