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

    Neuronpedia

    INDEX
    Explanations

    relationships

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    يل
    0.64
    了一個
    0.64
    }}{
    0.62
    ará
    0.62
     amerikan
    0.60
     моделей
    0.60
    heid
    0.59
     uitgevoerd
    0.59
    Е
    0.59
     élég
    0.58
    POSITIVE LOGITS
     friendships
    1.12
     relationships
    0.99
     relationship
    0.92
     camaraderie
    0.84
     friendship
    0.80
     relations
    0.79
     ties
    0.75
    誼
    0.73
    糀
    0.71
    relationships
    0.71
    Act Density 0.346%

    No Known Activations