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

    Neuronpedia

    INDEX
    Explanations

    Unfortunately, followed by a negative consequence

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    م
    2.80
    o
    2.66
    ר
    2.56
    ה
    2.50
    м
    2.17
    个
    2.14
    ない
    2.03
    ر
    2.02
    ளில்
    1.92
    er
    1.88
    POSITIVE LOGITS
     sigui
    1.70
    Jest
    1.66
     superconduct
    1.60
    倛
    1.60
     Retina
    1.59
     disadvant
    1.58
     Ім
    1.58
     regalos
    1.57
     puestos
    1.56
     modernize
    1.56
    Act Density 0.001%

    No Known Activations