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

    Neuronpedia

    INDEX
    Explanations

    have been or have seen

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     favourable
    0.81
    avorable
    0.80
     helpful
    0.77
     favorable
    0.74
    helpful
    0.74
     beneficial
    0.73
     Significant
    0.72
     significant
    0.71
    significant
    0.70
     ವಿಶೇಷ
    0.69
    POSITIVE LOGITS
     deserve
    1.02
     deserves
    0.95
     epit
    0.86
     belong
    0.83
     deserved
    0.79
     belonged
    0.79
     represents
    0.78
     embodies
    0.76
     promised
    0.76
     belongs
    0.75
    Act Density 0.265%

    No Known Activations