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

    Neuronpedia

    INDEX
    Explanations

    a lot of, difficult, following, help

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     lecz
    0.66
     असल्यास
    0.66
     אך
    0.65
     stets
    0.61
     آنان
    0.59
    Após
    0.58
     Após
    0.56
     НЕ
    0.53
     folgte
    0.52
    Jika
    0.52
    POSITIVE LOGITS
     ahorita
    1.06
     really
    1.03
     ähm
    1.01
     굉장히
    1.00
    比如说
    0.98
    いろんな
    0.96
     poquito
    0.95
     uh
    0.94
    っていう
    0.93
     väldigt
    0.93
    Act Density 0.017%

    No Known Activations