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

    Neuronpedia

    INDEX
    Explanations

    harmful requests

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     difficultés
    0.87
     différences
    0.86
     жүктөп
    0.85
     ennemis
    0.79
     erreurs
    0.79
    】，
    0.79
     inférieures
    0.78
     problèmes
    0.78
     migraines
    0.78
     mauvaise
    0.77
    POSITIVE LOGITS
     as
    1.09
     A
    1.03
     a
    1.01
     by
    0.94
    ,
    0.93
    .
    0.88
    A
    0.87
     c
    0.84
     s
    0.83
    1
    0.83
    Act Density 0.001%

    No Known Activations