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

    Neuronpedia

    INDEX
    Explanations

    December

    np_max-act · gemini-2.0-flash

    The neuron selectively activates on numeric tokens—especially decimal numbers (e.g. years or measurements with fractional parts).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     çağ
    -0.07
    signin
    -0.07
    ＆
    -0.07
     sizeof
    -0.07
    DataManager
    -0.07
    _flutter
    -0.06
    陣
    -0.06
     priest
    -0.06
    AntiForgeryToken
    -0.06
     copyrights
    -0.06
    POSITIVE LOGITS
    [channel
    0.07
    )↵↵↵↵↵↵
    0.06
     radial
    0.06
     framed
    0.06
    '};↵
    0.06
     Binder
    0.06
     JUL
    0.06
     }↵↵↵↵↵
    0.06
    			↵↵
    0.05
     Seeing
    0.05
    Act Density 0.006%

    No Known Activations