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

    Neuronpedia

    INDEX
    Explanations

    errors and crashes

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     निहित
    0.77
     renforcer
    0.75
     fortalecer
    0.74
     pasión
    0.72
     добы
    0.72
    simile
    0.72
     oficio
    0.71
     облег
    0.71
     fomentar
    0.70
     удобно
    0.70
    POSITIVE LOGITS
     crashes
    2.14
     crashing
    1.88
     crashed
    1.88
     crash
    1.87
     errors
    1.67
     malfunctions
    1.60
     failures
    1.57
     malfunction
    1.55
     glitches
    1.55
    crash
    1.54
    Act Density 1.210%

    No Known Activations