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

    Neuronpedia

    INDEX
    Explanations

    specifying use context

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     unrealistic
    0.62
     Š
    0.61
     определён
    0.61
     있지만
    0.60
     όλα
    0.60
     časti
    0.60
    ありますが
    0.60
     закончи
    0.60
    వలం
    0.60
     некоторые
    0.59
    POSITIVE LOGITS
    is
    0.85
    ش
    0.80
    on
    0.76
    ua
    0.76
    ud
    0.74
    ap
    0.74
    지
    0.70
    show
    0.68
    isierung
    0.68
    सिंग
    0.68
    Act Density 0.005%

    No Known Activations