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

    Neuronpedia

    INDEX
    Explanations

    actions and goals

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     vapors
    0.57
     stabilization
    0.57
     vaporization
    0.56
     fervor
    0.52
     caramelized
    0.51
     Всё
    0.50
     ধর্মীয়
    0.50
     stabilizing
    0.49
    fueled
    0.49
     Fue
    0.48
    POSITIVE LOGITS
     maximise
    1.15
     minimise
    0.98
     prioritise
    0.98
     utilise
    0.98
     utilises
    0.97
     realise
    0.96
     optimise
    0.95
     recognise
    0.95
     firstly
    0.94
     authorise
    0.92
    Act Density 0.008%

    No Known Activations