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

    Neuronpedia

    INDEX
    Explanations

    stop or crash

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     потери
    0.46
     ক্ষয়
    0.41
    蝕
    0.41
     loosing
    0.38
     Loss
    0.38
     Happened
    0.38
     arisen
    0.38
    ngthen
    0.38
    Loss
    0.38
     আবির্ভাব
    0.38
    POSITIVE LOGITS
     cease
    0.92
     ceases
    0.82
     ceased
    0.77
     completely
    0.74
     Ce
    0.70
     ce
    0.66
     permanently
    0.64
    Ce
    0.63
     simply
    0.59
     outright
    0.59
    Act Density 0.126%

    No Known Activations