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

    Neuronpedia

    INDEX
    Explanations

    across

    np_max-act-logits · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     funny
    0.50
    Funny
    0.49
     Funny
    0.48
    funny
    0.47
    ovno
    0.41
     humorous
    0.39
    赅
    0.38
     Colorful
    0.38
    dio
    0.38
     hilarious
    0.37
    POSITIVE LOGITS
     subject
    0.52
     wrongful
    0.45
     live
    0.44
    错误
    0.42
     incorrectly
    0.42
     defend
    0.40
     Subject
    0.40
     later
    0.38
    錯誤
    0.38
     invalidate
    0.38
    Act Density 0.001%

    No Known Activations