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

    Neuronpedia

    INDEX
    Explanations

    sexual coercion and safety violations

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ran
    0.50
    tab
    0.49
    tin
    0.48
     as
    0.48
    uran
    0.48
    ven
    0.46
    urin
    0.46
    aban
    0.46
    dan
    0.45
    as
    0.45
    POSITIVE LOGITS
    甃
    0.46
    ドラマ
    0.45
    ﻠ
    0.45
     glo
    0.43
    ﻔ
    0.43
     aest
    0.42
     ടീ
    0.42
     sống
    0.41
     beaux
    0.41
    罨
    0.41
    Act Density 0.003%

    No Known Activations