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

    Neuronpedia

    INDEX
    Explanations

    separating into parts

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Kato
    -0.74
     результатов
    -0.73
     workplaces
    -0.69
    𝐭
    -0.68
     extractor
    -0.66
    ornadas
    -0.64
    woke
    -0.64
     skirting
    -0.64
    skipped
    -0.63
     tune
    -0.63
    POSITIVE LOGITS
     Otis
    0.76
    ENOT
    0.72
    deville
    0.71
    MN
    0.70
    MSG
    0.70
     koop
    0.70
     حوالي
    0.69
     ГОСТ
    0.69
    BCD
    0.68
    AddRef
    0.67
    Act Density 0.066%

    No Known Activations