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

    Neuronpedia

    INDEX
    Explanations

    2

    np_max-act · gemini-2.0-flash

    This neuron flags numeric tokens—integers or decimal numbers—used in chemical names, measurements, or data values.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    liche
    -0.08
     хто
    -0.07
    irical
    -0.07
     WHICH
    -0.07
     Winds
    -0.07
     cables
    -0.06
    ického
    -0.06
    ými
    -0.06
    ző
    -0.06
     التح
    -0.06
    POSITIVE LOGITS
    >>();↵
    0.07
    war
    0.07
    ۲۵
    0.06
    !↵↵↵
    0.06
     ];↵↵
    0.06
    .Decode
    0.06
     $
↵
    0.06
    ]]);↵
    0.06
     Martian
    0.06
     деревян
    0.06
    Act Density 0.011%

    No Known Activations