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

    Neuronpedia

    INDEX
    Explanations

    pool of

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ный
    1.29
    ного
    1.26
    ко
    1.20
    is
    1.16
    не
    1.14
    vý
    1.02
    ри
    1.01
     شي
    0.96
    вторых
    0.96
     في
    0.95
    POSITIVE LOGITS
     of
    1.62
    ために
    1.51
     Pool
    1.35
    ל
    1.32
     pools
    1.30
    ન
    1.28
     at
    1.27
    ना
    1.27
     that
    1.24
    א
    1.23
    Act Density 0.005%

    No Known Activations