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

    Neuronpedia

    INDEX
    Explanations

    describing or troubleshooting a behavior or issue

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     failings
    0.45
     deficits
    0.43
     weaknesses
    0.42
     flaws
    0.41
     causas
    0.40
     inevitable
    0.40
     causes
    0.39
     inev
    0.39
     fraude
    0.39
     caused
    0.39
    POSITIVE LOGITS
     behavior
    0.88
     behaviour
    0.87
    behaviour
    0.82
     behav
    0.79
    行为
    0.77
     Behavior
    0.76
    behavior
    0.75
     Behaviour
    0.71
     поведения
    0.70
    Behavior
    0.68
    Act Density 0.034%

    No Known Activations