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

    Neuronpedia

    INDEX
    Explanations

    absolutely cannot

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     troughs
    0.54
     magari
    0.52
     inclinations
    0.50
     earners
    0.49
     victimization
    0.49
     flaps
    0.48
     dealings
    0.48
     elucidation
    0.48
    Usage
    0.48
    IFICATION
    0.47
    POSITIVE LOGITS
    very
    0.77
     absolutely
    0.70
    absolutely
    0.68
     VERY
    0.67
     very
    0.67
     খুবই
    0.63
     velmi
    0.63
     بسیار
    0.62
     очень
    0.61
    only
    0.61
    Act Density 0.366%

    No Known Activations