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

    Neuronpedia

    INDEX
    Explanations

    separable equations

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
     constitu
    -0.08
    erdem
    -0.07
     Seating
    -0.07
    -0.07
     winners
    -0.07
     السيا
    -0.07
     suspicious
    -0.07
     illumination
    -0.07
     제공
    -0.07
     winning
    -0.07
    POSITIVE LOGITS
     drank
    0.10
    _drag
    0.09
     կանխ
    0.09
     nonlinear
    0.09
     propelled
    0.09
     governed
    0.09
    .proto
    0.08
     Fahrt
    0.08
    Dragged
    0.08
     հեռ
    0.08
    Act Density 0.012%
    No Known Activations