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

    Neuronpedia

    INDEX
    Explanations

    programming classes

    np_max-act · gemini-2.0-flash

    The neuron is chiefly detecting floating‐point numeric literals (i.e. decimal numbers).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    _AUTHOR
    -0.07
    :r
    -0.07
     ocean
    -0.07
    <Book
    -0.07
    (Parameter
    -0.07
     bý
    -0.06
    subscribe
    -0.06
     Prim
    -0.06
    (cube
    -0.06
    说
    -0.06
    POSITIVE LOGITS
    -port
    0.06
     '
↵
    0.06
     HCI
    0.06
    यह
    0.06
     UCLA
    0.06
     양
    0.05
    _MAJOR
    0.05
    awy
    0.05
    (cluster
    0.05
    нося
    0.05
    Act Density 0.330%

    No Known Activations