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

    Neuronpedia

    INDEX
    Explanations

    words preceding punctuation

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     slow
    0.60
     afl
    0.60
     impervious
    0.55
    传递
    0.54
     kq
    0.54
     $|
    0.52
     fear
    0.51
    *;
    0.51
     共
    0.51
     regulated
    0.51
    POSITIVE LOGITS
    HE
    0.77
    YOU
    0.75
    WHAT
    0.72
    EEN
    0.72
    THERE
    0.71
    SOME
    0.70
     عَل
    0.68
    EL
    0.68
     aquell
    0.68
    Wouldn
    0.68
    Act Density 0.041%

    No Known Activations