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

    Neuronpedia

    INDEX
    Explanations

    setting options or properties

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    на
    1.10
     chimiques
    1.02
    ك
    1.01
     jede
    0.92
     inz
    0.90
    ка
    0.89
    ми
    0.87
    ستي
    0.86
    在
    0.86
     في
    0.84
    POSITIVE LOGITS
    ’
    1.38
    o
    1.27
    on
    1.20
    is
    1.08
    et
    1.02
    u
    1.00
     It
    0.97
    ↵↵
    0.96
    as
    0.95
    िंग्स
    0.94
    Act Density 0.039%

    No Known Activations