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

    Neuronpedia

    INDEX
    Explanations

    In this blog post, we will discuss

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    '
    1.22
    лно
    1.14
    лни
    1.08
    เนื่อง
    1.04
    $',
    1.03
    ushing
    1.02
    kyverno
    0.96
    δήποτε
    0.95
    *',
    0.92
    _'
    0.91
    POSITIVE LOGITS
    ى
    1.34
     appease
    1.09
    c
    1.09
    or
    1.01
     Cliquez
    0.97
    ка
    0.93
     replace
    0.91
    ა
    0.89
     appelée
    0.89
    ja
    0.88
    Act Density 0.013%

    No Known Activations