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

    Neuronpedia

    INDEX
    Explanations

    but rather

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     dunque
    0.55
     więc
    0.52
     donc
    0.50
    ົບ
    0.50
    จึง
    0.49
     جعلی
    0.49
    조
    0.47
     つ
    0.47
     daher
    0.47
     quindi
    0.47
    POSITIVE LOGITS
    Nevertheless
    0.57
    termilk
    0.57
     dennoch
    0.55
     nevertheless
    0.52
     nonetheless
    0.52
    ted
    0.49
    But
    0.49
    thole
    0.48
     Nevertheless
    0.46
     మాత్రం
    0.46
    Act Density 0.012%

    No Known Activations