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

    Neuronpedia

    INDEX
    Explanations

    struggle to achieve actions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     ermöglicht
    0.55
     lehető
    0.50
    hopefully
    0.49
     hopefully
    0.47
     allows
    0.46
     позволяет
    0.45
     ermöglichen
    0.44
     Hopefully
    0.44
     omoguć
    0.44
     যাতে
    0.43
    POSITIVE LOGITS
     siquiera
    0.88
     anymore
    0.76
     unless
    0.74
     anything
    0.73
    unless
    0.71
    任何
    0.70
    anything
    0.70
    Unless
    0.61
     ვერ
    0.60
     qualquer
    0.60
    Act Density 0.157%

    No Known Activations