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

    Neuronpedia

    INDEX
    Explanations

    parenthetical remarks and emojis

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     essential
    0.86
     ajust
    0.82
     of
    0.82
     on
    0.80
     podia
    0.80
     instrumental
    0.77
     under
    0.76
     includ
    0.75
     basis
    0.75
     integral
    0.74
    POSITIVE LOGITS
    Honestly
    0.97
    ل
    0.92
    ת
    0.90
    Ironically
    0.87
    न
    0.87
     Aunque
    0.86
    ان
    0.85
    Actually
    0.85
    Didn
    0.85
     Даже
    0.84
    Act Density 0.000%

    No Known Activations