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

    Neuronpedia

    INDEX
    Explanations

    LLaMA model

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    is
    0.57
    high
    0.50
    RE
    0.49
     astronauts
    0.49
    ↵
    0.49
     troops
    0.48
     alta
    0.48
     R
    0.48
    REN
    0.48
    ün
    0.47
    POSITIVE LOGITS
     подробно
    0.50
    伷
    0.48
     проте
    0.48
    язы
    0.47
     insofar
    0.47
    SubDirectoryPath
    0.46
    文學
    0.46
     inasmuch
    0.46
     бухгалтер
    0.45
    混乱
    0.44
    Act Density 0.005%

    No Known Activations