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

    Neuronpedia

    INDEX
    Explanations

    key checking, volume data

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    felter
    0.69
     substantiated
    0.68
     (’
    0.68
    정을
    0.66
    вид
    0.66
    ě
    0.66
    textepsilon
    0.64
    ول
    0.63
    ారు
    0.63
    ющий
    0.62
    POSITIVE LOGITS
    ↵
    0.94
    는
    0.93
    ing
    0.88
    ด
    0.88
    是
    0.84
    は
    0.84
    k
    0.82
    v
    0.81
    is
    0.79
    be
    0.78
    Act Density 0.000%

    No Known Activations