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

    Neuronpedia

    INDEX
    Explanations

    never to always

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     everything
    0.82
    everything
    0.73
     덕
    0.71
    的东西
    0.69
     noastră
    0.68
     quantos
    0.67
    到时候
    0.66
    まさに
    0.65
    कायदा
    0.64
     ведь
    0.63
    POSITIVE LOGITS
    Never
    1.69
    Always
    1.64
     Never
    1.62
     Always
    1.60
     never
    1.55
     NEVER
    1.50
     always
    1.47
    never
    1.46
    always
    1.46
     ALWAYS
    1.44
    Act Density 0.176%

    No Known Activations