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

    Neuronpedia

    INDEX
    Explanations

    python function definitions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     is
    0.40
     was
    0.31
    '
    0.29
    溂
    0.29
     sangue
    0.27
     vanishes
    0.27
    ’
    0.27
     brasileiro
    0.26
     accuses
    0.26
    ação
    0.25
    POSITIVE LOGITS
    ad
    0.32
    modern
    0.32
    ل
    0.32
    extremely
    0.30
    f
    0.30
    л
    0.30
    ре
    0.29
    р
    0.29
    ർ
    0.29
    short
    0.29
    Act Density 1.214%

    No Known Activations