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

    Neuronpedia

    INDEX
    Explanations

    highlighting, truthful, short, improve, converting

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Shields
    0.38
     Shield
    0.38
    Sheet
    0.38
     shield
    0.36
    Shield
    0.35
    tywn
    0.35
     shields
    0.35
     weakened
    0.35
    ologis
    0.34
    圣
    0.34
    POSITIVE LOGITS
     کاب
    0.43
    inoza
    0.41
     cabinets
    0.40
    ransition
    0.39
     बुल
    0.39
     Ste
    0.39
    厸
    0.39
    lemagne
    0.38
     escalator
    0.38
     hemicontinuous
    0.38
    Act Density 0.000%

    No Known Activations