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

    Neuronpedia

    INDEX
    Explanations

    unexpected conjunctions, yet

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     많이
    0.40
     lagen
    0.39
    Truly
    0.39
     보시면
    0.38
     raczej
    0.38
     desir
    0.37
    ৃতি
    0.36
     Differ
    0.36
    不太
    0.35
    Lx
    0.35
    POSITIVE LOGITS
    居然
    1.30
     suddenly
    1.23
    竟然
    1.15
     plötzlich
    1.02
     вдруг
    1.01
    なのに
    0.98
     Suddenly
    0.94
     inexplic
    0.93
     pourtant
    0.88
    忽然
    0.83
    Act Density 0.019%

    No Known Activations