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

    Neuronpedia

    INDEX
    Explanations

    index followed by parenthesis

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    屃
    -1.70
    腘
    -1.60
     což
    -1.59
    ኧ
    -1.52
    騭
    -1.42
     protože
    -1.41
     dicho
    -1.39
    </h1>
    -1.38
    鐿
    -1.38
    我知道
    -1.37
    POSITIVE LOGITS
    er
    1.84
     at
    1.60
     now
    1.52
     most
    1.51
     still
    1.49
    ists
    1.45
    for
    1.41
     on
    1.41
    .
    1.39
     слегка
    1.38
    Act Density 0.008%

    No Known Activations