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

    Neuronpedia

    INDEX
    Explanations

    吃 话 计划 看看 烦

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    縝
    0.69
    蟄
    0.67
    鲽
    0.65
    تباين
    0.63
    殞
    0.63
    繚
    0.62
    皚
    0.61
    堊
    0.61
    鷗
    0.61
    規範
    0.60
    POSITIVE LOGITS
    吃
    0.82
    做
    0.70
    没
    0.70
     scolded
    0.69
    打
    0.69
    怎么
    0.68
    不好
    0.68
    买
    0.68
    老师
    0.66
     吃
    0.66
    Act Density 0.011%

    No Known Activations