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

    Neuronpedia

    INDEX
    Explanations

    breakdown or explanation covering

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     faintly
    0.42
     ఏద
    0.39
    ientôt
    0.39
     sélectionnés
    0.38
     inventing
    0.37
    стья
    0.36
     ইচ্ছে
    0.36
     অসম্ভব
    0.35
     Inserting
    0.35
     Select
    0.35
    POSITIVE LOGITS
     breakdown
    2.05
     break
    2.03
    Break
    1.90
     breaking
    1.88
     broken
    1.88
     Breakdown
    1.86
    break
    1.84
     Break
    1.83
     breakdowns
    1.80
     breaks
    1.79
    Act Density 0.703%

    No Known Activations