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

    Neuronpedia

    INDEX
    Explanations

    survey results

    np_max-act · gemini-2.0-flash

    The neuron never actually fires on any of the provided text—it does not respond to words, punctuation, or formatting—but would in principle be looking for numeric tokens (e.g. the floating‐point score values) rather than ordinary words.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    حد
    -0.08
     OL
    -0.06
    ICollectionView
    -0.06
    handlers
    -0.06
     boxed
    -0.06
    446
    -0.06
    podob
    -0.06
    icios
    -0.06
    體
    -0.06
    urança
    -0.06
    POSITIVE LOGITS
     Nation
    0.07
    \\"
    0.06
     intersects
    0.06
    contents
    0.06
    でも
    0.06
    .iso
    0.06
     Viet
    0.06
    beans
    0.06
     Download
    0.06
    او
    0.06
    Act Density 0.005%

    No Known Activations