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

    Neuronpedia

    INDEX
    Explanations

    leveling, pooling, rollback, distillation

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ⚕
    0.47
     χωρίς
    0.46
     interstitiis
    0.45
     ombil
    0.45
    琚
    0.45
     WLAN
    0.44
    ากาศ
    0.43
     âng
    0.43
    <unused104>
    0.43
     Wix
    0.42
    POSITIVE LOGITS
    (
    0.54
     حالة
    0.45
    크
    0.44
    （
    0.43
    ↵↵
    0.42
    +
    0.42
    ز
    0.41
    <0xE8>
    0.41
    פו
    0.41
    터
    0.40
    Act Density 0.373%

    No Known Activations