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

    Neuronpedia

    INDEX
    Explanations

    Exp

    np_max-act · gemini-2.0-flash

    The neuron strongly activates on numeric tokens representing decimal fractions (floating-point numbers).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     factions
    -0.07
     Te
    -0.07
    鹿
    -0.07
     Amir
    -0.06
    _prices
    -0.06
     ("/
    -0.06
     Grammar
    -0.06
    Wrong
    -0.06
    .selectAll
    -0.06
     Vaults
    -0.06
    POSITIVE LOGITS
     İnsan
    0.06
    �
    0.06
    Toyota
    0.06
    raj
    0.06
     scrolled
    0.06
     sữa
    0.06
    刚
    0.06
    ynamo
    0.06
     چگونه
    0.06
    (exc
    0.06
    Act Density 0.002%

    No Known Activations