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

    Neuronpedia

    INDEX
    Explanations

    or

    np_max-act · gemini-2.0-flash

    The neuron strongly activates on numeric tokens and related date or ordinal expressions (e.g. digits, “th,” month names, etc.).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    Compatibility
    -0.07
     Extraction
    -0.07
     Thánh
    -0.07
     Mei
    -0.07
     Flynn
    -0.06
    Demon
    -0.06
    _BACKGROUND
    -0.06
    _Min
    -0.06
     anne
    -0.06
    Matt
    -0.06
    POSITIVE LOGITS
    .JLabel
    0.06
    year
    0.06
     juegos
    0.06
     advises
    0.06
    _all
    0.06
     Appalach
    0.06
    imet
    0.05
    jab
    0.05
    龙
    0.05
    productId
    0.05
    Act Density 0.111%

    No Known Activations