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

    Neuronpedia

    INDEX
    Explanations

    over and thousands of

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     be
    0.49
     een
    0.47
     to
    0.45
     として
    0.42
     ért
    0.39
     ç
    0.38
     ein
    0.38
     impegno
    0.36
     is
    0.36
     で
    0.36
    POSITIVE LOGITS
     Globally
    0.34
    Tutorial
    0.31
    The
    0.31
    a
    0.31
     The
    0.30
    Hawaii
    0.30
    로는
    0.29
     Tutorials
    0.28
     fewer
    0.28
    린
    0.28
    Act Density 0.026%

    No Known Activations