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

    Neuronpedia

    INDEX
    Explanations

    unexpected outcome or realization

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    会导致
    0.49
    ｛
    0.39
     extraer
    0.39
     указывает
    0.39
     determin
    0.38
     mappings
    0.38
     denotes
    0.38
     infertility
    0.38
    ₚ
    0.38
     hypothalamus
    0.38
    POSITIVE LOGITS
    果然
    0.65
     оказалось
    0.64
     okaza
    0.61
     pleasantly
    0.56
     proved
    0.55
     оказалась
    0.52
     surprisingly
    0.52
    surprisingly
    0.50
     оказался
    0.49
     Indeed
    0.49
    Act Density 0.041%

    No Known Activations