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

    Neuronpedia

    INDEX
    Explanations

    trained

    np_max-act-logits · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Programming
    0.66
    programming
    0.64
     programming
    0.63
     educators
    0.62
     принципе
    0.60
    लेज
    0.59
     artillery
    0.59
     consulting
    0.58
    iply
    0.58
    চেতন
    0.57
    POSITIVE LOGITS
     Upon
    1.08
    Upon
    1.03
     sobre
    1.03
    上面
    1.00
     upon
    0.99
     върху
    0.93
    上了
    0.93
    上的
    0.90
     atas
    0.90
    upon
    0.87
    Act Density 0.015%

    No Known Activations