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

    Neuronpedia

    INDEX
    Explanations

    usual occurrence

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     typically
    1.15
     Typically
    1.10
    Typically
    0.95
    typically
    0.89
     typical
    0.85
     típico
    0.83
     particularmente
    0.80
     típica
    0.76
     típ
    0.73
     particular
    0.72
    POSITIVE LOGITS
    verhalten
    0.47
    sofar
    0.46
    Found
    0.46
    就好
    0.45
     esetben
    0.45
    Sort
    0.44
    idung
    0.44
    遇到的
    0.44
    scenario
    0.44
    做法
    0.43
    Act Density 0.034%

    No Known Activations