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

    Neuronpedia

    INDEX
    Explanations

    multiply each

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    str
    0.42
    0.41
    reek
    0.41
    fect
    0.40
     Стро
    0.40
    wa
    0.40
    ärm
    0.40
    amanan
    0.40
    ारों
    0.40
     최고의
    0.40
    POSITIVE LOGITS
     liber
    0.49
     impoverished
    0.43
     Liber
    0.42
     légèrement
    0.42
     modu
    0.42
     displacements
    0.41
     skon
    0.40
     roug
    0.40
     cartil
    0.40
     alumina
    0.40
    Act Density 0.010%

    No Known Activations