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

    Neuronpedia

    INDEX
    Explanations

    better, more likely

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     somewhat
    0.66
     agak
    0.61
     slightly
    0.56
    有點
    0.52
     trochu
    0.52
     strict
    0.51
     strictly
    0.50
     sometimes
    0.50
     약간
    0.49
     nieco
    0.48
    POSITIVE LOGITS
     легче
    0.82
    更容易
    0.80
     easier
    0.71
     happier
    0.70
     Easier
    0.63
     healthier
    0.60
    更好地
    0.59
     bättre
    0.58
     καλύτε
    0.57
     лучше
    0.56
    Act Density 0.015%

    No Known Activations