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

    Neuronpedia

    INDEX
    Explanations

    i'm sorry / i'm going

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Used
    0.87
    Typically
    0.80
     wanting
    0.76
     mêmes
    0.75
    Similar
    0.75
    enjoy
    0.74
    Usually
    0.71
     able
    0.71
    typically
    0.70
     typically
    0.70
    POSITIVE LOGITS
     going
    1.10
     सॉरी
    1.06
     sorry
    1.05
     Going
    1.01
    GOING
    0.97
     vont
    0.95
     gonna
    0.94
    Going
    0.90
     goin
    0.89
     Sorry
    0.86
    Act Density 0.053%

    No Known Activations