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

    Neuronpedia

    INDEX
    Explanations

    joke setup "Why

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     initio
    0.77
    thumb
    0.77
    Thumb
    0.72
     Coldplay
    0.71
     crumbs
    0.70
     কাশী
    0.68
     órgão
    0.68
     suicidal
    0.68
     spores
    0.67
     proverbial
    0.67
    POSITIVE LOGITS
    呼び
    0.77
    走
    0.76
     идти
    0.76
     acompaña
    0.74
    Walking
    0.74
     busy
    0.73
    Coming
    0.73
     walked
    0.73
     acompan
    0.72
     Walking
    0.72
    Act Density 0.052%

    No Known Activations