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

    Neuronpedia

    INDEX
    Explanations

    responds to changes and stimuli

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     responsible
    0.66
     caused
    0.66
    unsafe
    0.62
     relying
    0.59
     resorting
    0.58
     risking
    0.58
    ing
    0.58
    resulting
    0.58
    causing
    0.57
    fired
    0.57
    POSITIVE LOGITS
    |,
    0.98
     extranjeros
    0.85
     अगदी
    0.84
     quase
    0.84
     бонусу
    0.81
     kahit
    0.81
     criticism
    0.80
     sfera
    0.80
     vVertex
    0.79
     tempData
    0.79
    Act Density 0.104%

    No Known Activations