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

    Neuronpedia

    INDEX
    Explanations

    past echoes and memories

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     unusual
    0.77
     new
    0.73
    cie
    0.71
     nye
    0.71
    INCREMENT
    0.68
     unconventional
    0.67
    ेटेड
    0.66
     visibility
    0.66
     novo
    0.65
    lowing
    0.64
    POSITIVE LOGITS
     bygone
    2.02
     прошлого
    1.55
     past
    1.55
    époque
    1.53
     과거
    1.51
     преж
    1.46
     época
    1.43
     Vergangenheit
    1.43
     yester
    1.42
     memories
    1.39
    Act Density 0.141%

    No Known Activations