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

    Neuronpedia

    INDEX
    Explanations

    template of a problem definition

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     ☐
    0.44
    آم
    0.43
     Architect
    0.42
     architect
    0.42
    architect
    0.41
     mpg
    0.41
     coordinators
    0.41
    クシー
    0.41
    dozen
    0.41
    shell
    0.40
    POSITIVE LOGITS
    不断
    0.45
    ן
    0.45
    不斷
    0.45
     函数
    0.44
     Neuer
    0.44
    烧
    0.44
    变化
    0.43
    ץ
    0.43
     עד
    0.43
    钝
    0.43
    Act Density 0.001%

    No Known Activations