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

    Neuronpedia

    INDEX
    Explanations

    is determined by

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Which
    0.78
     Такие
    0.75
     които
    0.73
    Such
    0.72
     Which
    0.70
    которые
    0.66
     které
    0.63
     niektórych
    0.63
     Such
    0.61
    któ
    0.61
    POSITIVE LOGITS
     driven
    1.23
     rooted
    1.22
     simply
    1.20
     about
    1.20
     akin
    1.15
     merely
    1.15
     based
    1.13
     indicative
    1.12
     comprised
    1.11
     dictated
    1.10
    Act Density 0.894%

    No Known Activations