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

    Neuronpedia

    INDEX
    Explanations

    words can't express

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    動作
    0.40
     exception
    0.38
     maintained
    0.38
     unstable
    0.38
     frowned
    0.38
     dairy
    0.38
     vegetable
    0.38
     disregarding
    0.38
     neglected
    0.37
     disregarded
    0.37
    POSITIVE LOGITS
     comprehension
    0.66
    adequ
    0.65
     comprehend
    0.64
     adequately
    0.63
     Compre
    0.61
     compreender
    0.59
     comprendere
    0.58
     शब्दों
    0.57
     comprender
    0.57
     Words
    0.56
    Act Density 0.010%

    No Known Activations