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

    Neuronpedia

    INDEX
    Explanations

    a key component

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Every
    0.47
    ровка
    0.46
     отрима
    0.45
     நான்கு
    0.45
     mudança
    0.44
     наші
    0.44
     можуть
    0.44
     refusal
    0.43
     monstros
    0.43
    0.43
    POSITIVE LOGITS
     bese
    0.38
     guna
    0.37
     parallel
    0.37
     attached
    0.36
    yl
    0.35
     poll
    0.35
     Schlü
    0.35
    speaker
    0.34
     part
    0.34
    deceased
    0.34
    Act Density 0.019%

    No Known Activations