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

    Neuronpedia

    INDEX
    Explanations

    pauses and interjections

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ,
    1.14
    ，
    1.06
    ،
    0.99
    、
    0.86
    (),
    0.86
    ],
    0.84
    那种
    0.82
    "،
    0.82
    .$,
    0.79
    [{
    0.78
    POSITIVE LOGITS
     ờ
    0.93
     ehm
    0.90
     пожалуйста
    0.86
     uh
    0.84
     well
    0.81
     respectiv
    0.75
     hogy
    0.72
    enia
    0.72
     bhikkhave
    0.70
     بدور
    0.69
    Act Density 0.095%

    No Known Activations