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

    Neuronpedia

    INDEX
    Explanations

    pg packages and extensions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Herb
    0.64
     Me
    0.63
     Do
    0.60
     Qual
    0.60
     boldness
    0.59
     certainty
    0.59
     shoulder
    0.58
     Acids
    0.58
     Cur
    0.57
     Cast
    0.57
    POSITIVE LOGITS
    开源
    1.04
    软件包
    1.03
    [multimodal]
    1.00
    命令行
    0.95
    plotlib
    0.95
    来实现
    0.94
    elizmente
    0.91
    库
    0.91
     および
    0.89
    模块
    0.88
    Act Density 1.379%

    No Known Activations