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

    Neuronpedia

    INDEX
    Explanations

    friendship and emotional bonds

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     starve
    0.66
     softmax
    0.65
    utf
    0.65
     ReLU
    0.64
    納期
    0.63
    oxins
    0.63
     乾燥
    0.63
     işlemler
    0.62
     chcesz
    0.62
    찝
    0.62
    POSITIVE LOGITS
     bond
    1.42
     bonds
    1.31
     camaraderie
    1.23
     friendship
    1.19
     friendships
    1.17
     bonded
    1.09
     companionship
    1.07
     Bond
    1.06
     loyal
    1.04
     special
    1.03
    Act Density 0.172%

    No Known Activations