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

    Neuronpedia

    INDEX
    Explanations

    end of turn markers

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    are
    1.37
    いない
    1.13
    ot
    1.10
    ج
    1.10
    1
    1.08
     evit
    1.05
    違う
    1.03
    int
    1.02
    ignores
    1.02
     bashing
    1.02
    POSITIVE LOGITS
     کدام
    1.12
    отя
    1.05
    лла
    1.04
    таў
    1.03
     Люди
    1.02
     therapeut
    1.01
     Ủy
    0.99
     σημ
    0.97
    icación
    0.95
    gruppe
    0.94
    Act Density 0.001%

    No Known Activations