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

    Neuronpedia

    INDEX
    Explanations

    lists of ways or reasons

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     two
    0.67
     dwóch
    0.64
     zwei
    0.64
     deux
    0.63
     രണ്ടു
    0.60
     δύο
    0.58
     două
    0.57
     двох
    0.57
     iki
    0.55
     двумя
    0.55
    POSITIVE LOGITS
     Ways
    0.78
     Reasons
    0.75
     ways
    0.75
     reasons
    0.72
    questions
    0.64
     maneiras
    0.64
    Reasons
    0.63
    Ways
    0.62
    理由
    0.62
     examples
    0.61
    Act Density 0.009%

    No Known Activations