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

    Neuronpedia

    INDEX
    Explanations

    Pareto Principle (80/20 Rule)

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     koristiti
    0.47
     दिलों
    0.44
    цима
    0.44
    ávání
    0.43
    䥑
    0.43
    얏
    0.43
     센
    0.42
     모델
    0.42
     সন্ব
    0.42
    慷
    0.42
    POSITIVE LOGITS
     strained
    0.38
     American
    0.37
    स्थित
    0.37
    Mark
    0.36
     သည်
    0.36
     or
    0.35
     Russian
    0.35
     Mark
    0.35
    残り
    0.35
     remaining
    0.34
    Act Density 0.005%

    No Known Activations