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

    Neuronpedia

    INDEX
    Explanations

    code quality and correctness

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Ches
    0.87
     დროს
    0.82
    悧
    0.80
    пуляр
    0.80
     進
    0.80
     Lesser
    0.79
     عرص
    0.78
     علاقه
    0.78
    ೦
    0.77
     사이
    0.77
    POSITIVE LOGITS
     efficiency
    0.85
     efficient
    0.76
     complete
    0.75
     portability
    0.74
     stability
    0.71
     elegantly
    0.71
     elegant
    0.70
     completo
    0.70
     skipping
    0.70
     readability
    0.69
    Act Density 0.119%

    No Known Activations