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

    Neuronpedia

    INDEX
    Explanations

    acknowledging a point before contrasting

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    不仅
    0.57
     nejen
    0.57
     либо
    0.56
     either
    0.54
    Either
    0.49
     настолько
    0.45
     simplesmente
    0.44
    либо
    0.44
     entweder
    0.43
    不僅
    0.42
    POSITIVE LOGITS
     ostensibly
    0.94
     nominally
    0.93
     superficially
    0.86
     technically
    0.82
     certainly
    0.82
     Certainly
    0.79
     undoubtedly
    0.73
    確かに
    0.73
     admittedly
    0.72
     certamente
    0.71
    Act Density 0.097%

    No Known Activations