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

    Neuronpedia

    INDEX
    Explanations

    Equations

    np_max-act · gemini-2.0-flash

    The neuron responds to numeric constants (tokens consisting of digits, e.g. “143,” “-22,” “7,” “-359,” etc.).

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

    mathematical expressions and equations.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    ilation
    -0.07
    wap
    -0.06
     opens
    -0.06
    .dismiss
    -0.06
    PublicKey
    -0.06
     cập
    -0.06
    utter
    -0.06
     Peel
    -0.06
    <Texture
    -0.06
     Hamp
    -0.06
    POSITIVE LOGITS
     CONS
    0.06
    �
    0.06
    örper
    0.06
    OUN
    0.06
    rement
    0.06
     unw
    0.06
    952
    0.06
    $,
    0.05
    zung
    0.05
     greatly
    0.05
    Act Density 0.038%

    No Known Activations