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

    Neuronpedia

    INDEX
    Explanations

    state remains stable

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     cuyas
    0.54
     whose
    0.50
    whose
    0.50
     cuja
    0.50
    '
    0.49
     cuya
    0.47
     cujo
    0.47
     ktor
    0.47
    用於
    0.46
    사용
    0.45
    POSITIVE LOGITS
     varied
    0.71
     vary
    0.66
     proportional
    0.60
     varies
    0.60
     fluctuate
    0.59
     다양
    0.59
     அதிகரிக்க
    0.59
     malleable
    0.57
     محدود
    0.55
     commensurate
    0.55
    Act Density 0.017%

    No Known Activations