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

    Neuronpedia

    INDEX
    Explanations

    equality/comparison

    np_max-act-logits · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    หลัง
    0.72
     ♪
    0.65
     &(
    0.65
    ワールド
    0.61
    なんです
    0.59
    cein
    0.59
     glazed
    0.58
     Certainly
    0.58
     Multi
    0.57
     ******
    0.57
    POSITIVE LOGITS
     eq
    1.50
    eq
    1.46
    Eq
    1.34
     Eq
    1.29
     EQ
    1.27
     equ
    1.16
     eqs
    1.13
    EQ
    1.12
     ne
    1.06
     Eqs
    1.06
    Act Density 0.176%

    No Known Activations