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

    Neuronpedia

    INDEX
    Explanations

    ReLU activation

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
     concent
    -0.07
     deduct
    -0.07
    antage
    -0.07
     dishes
    -0.07
     maga
    -0.07
    Nr
    -0.07
    Reasons
    -0.07
    -host
    -0.07
    Compar
    -0.07
     Mun
    -0.07
    POSITIVE LOGITS
     flowing
    0.08
     ELSE
    0.08
     valent
    0.08
     slopes
    0.07
     behalve
    0.07
     leef
    0.07
     töl
    0.07
    erval
    0.07
    .Activity
    0.07
     harmless
    0.07
    Act Density 0.004%

    No Known Activations