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

    Neuronpedia

    INDEX
    Explanations

    quantifying capacity, length, width, and limits

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     names
    0.40
    стно
    0.40
    的名字
    0.39
    තා
    0.38
    F
    0.38
    Contra
    0.38
    G
    0.37
    ↵↵
    0.37
    HING
    0.37
    Citi
    0.36
    POSITIVE LOGITS
    是多少
    0.60
     limite
    0.60
     составляет
    0.56
    为
    0.55
     $=
    0.53
     wynosi
    0.51
     beträgt
    0.51
     limites
    0.51
    ＝
    0.50
     <=
    0.50
    Act Density 0.735%

    No Known Activations