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

    Neuronpedia

    INDEX
    Explanations

    averaging costs and frequencies

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     denial
    0.41
     unw
    0.41
     denied
    0.39
     U
    0.38
    反應
    0.38
     Denial
    0.37
     fio
    0.36
     forest
    0.36
    cut
    0.36
     interrupt
    0.36
    POSITIVE LOGITS
     occasionally
    0.53
    Occasionally
    0.48
     Occasionally
    0.47
     Часто
    0.45
     útil
    0.42
    偶尔
    0.42
    requently
    0.41
     часто
    0.41
    ATEC
    0.41
    ometimes
    0.40
    Act Density 0.003%

    No Known Activations