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

    Neuronpedia

    INDEX
    Explanations

    detecting quality issues

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     soigne
    0.46
    創作
    0.44
    操控
    0.41
     manipul
    0.41
     रचना
    0.39
     curving
    0.38
    fez
    0.38
    ្សែ
    0.38
    課
    0.38
    োজ
    0.38
    POSITIVE LOGITS
     detect
    1.11
    检测
    1.10
     Detection
    1.09
     detection
    1.08
     detects
    1.07
     detecting
    1.05
    Detect
    1.05
    Detection
    1.04
     Detecting
    1.03
     检测
    1.02
    Act Density 0.016%

    No Known Activations