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

    Neuronpedia

    INDEX
    Explanations

    conversion efficiency

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Nella
    0.42
    透过
    0.41
    ,
    0.40
     originalmente
    0.39
     origine
    0.39
    READY
    0.38
    烺
    0.38
    Re
    0.38
    Originally
    0.38
    Pac
    0.37
    POSITIVE LOGITS
     blames
    0.50
     केल्यानंतर
    0.48
     எடை
    0.47
     বউ
    0.47
     fallait
    0.45
     मजेदार
    0.45
     losers
    0.44
     {"
    0.44
    してください
    0.44
     pouvaient
    0.44
    Act Density 0.001%

    No Known Activations