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

    Neuronpedia

    INDEX
    Explanations

    concession and contrast

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Even
    0.48
    difficult
    0.48
    EVEN
    0.45
    残念
    0.43
    Unfortunately
    0.43
     어려
    0.42
    Sadly
    0.42
     EVEN
    0.41
     difficile
    0.41
    难以
    0.41
    POSITIVE LOGITS
     nevertheless
    1.30
     nonetheless
    1.28
     dennoch
    1.13
     Nevertheless
    1.06
    Nevertheless
    1.06
     trotzdem
    1.01
     Nonetheless
    0.96
    Nonetheless
    0.93
    それでも
    0.91
     néanmoins
    0.89
    Act Density 0.017%

    No Known Activations