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

    Neuronpedia

    INDEX
    Explanations

    use code blocks for formatting

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     dynam
    0.59
    Dynamic
    0.59
     dinam
    0.58
     динами
    0.58
     dynamic
    0.57
     Dynamic
    0.57
     Dynam
    0.57
    Dynam
    0.56
    dynamic
    0.55
    DYNAMIC
    0.54
    POSITIVE LOGITS
     code
    1.20
     Code
    1.05
    code
    1.03
     CODE
    0.97
    代码
    0.96
    Code
    0.91
     código
    0.90
     코드
    0.90
     код
    0.89
     कोड
    0.88
    Act Density 0.032%

    No Known Activations