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

    Neuronpedia

    INDEX
    Explanations

    transparency

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ra
    0.99
    ry
    0.95
    tait
    0.94
    tere
    0.88
    层次
    0.84
    la
    0.81
    ray
    0.80
    ul
    0.79
    leri
    0.78
    ram
    0.77
    POSITIVE LOGITS
     Э
    0.79
    щением
    0.77
     ES
    0.74
    НЫ
    0.73
     UTS
    0.73
     sinners
    0.72
    エース
    0.72
     ба
    0.72
     reigned
    0.71
     イ
    0.70
    Act Density 0.002%

    No Known Activations