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

    Neuronpedia

    INDEX
    Explanations

    universe size

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    is
    0.47
    ----------------
    0.46
    t
    0.46
    as
    0.45
    operatorname
    0.44
    na
    0.43
    aj
    0.42
     পাওয়া
    0.42
    s
    0.42
     introduced
    0.41
    POSITIVE LOGITS
     Hochzeit
    0.53
     kız
    0.47
    甃
    0.45
     कन्या
    0.44
    ంతరం
    0.43
     কন্যা
    0.43
    Scaling
    0.43
    ិត
    0.43
     klim
    0.42
    ereur
    0.42
    Act Density 0.006%

    No Known Activations