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

    Neuronpedia

    INDEX
    Explanations

    most frequent number or type

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ه
    1.77
    er
    1.25
     نیرو
    1.20
    ighet
    1.18
    ську
    1.17
    performs
    1.12
    assim
    1.12
     fratt
    1.12
    an
    1.11
    ה
    1.11
    POSITIVE LOGITS
    scriptstyle
    1.25
    ibly
    1.19
    ELSE
    1.11
     candied
    1.09
    ㄙ
    1.07
     variegated
    1.06
    tter
    1.06
     меры
    1.05
     triples
    1.05
     sealed
    1.04
    Act Density 0.000%

    No Known Activations