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

    Neuronpedia

    INDEX
    Explanations

    duplicated, stacked, workflows, without

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     A
    0.46
     wird
    0.46
     Projekte
    0.44
    できます
    0.44
     davon
    0.43
     മാത്രമേ
    0.43
    된다
    0.42
     becomes
    0.41
     Projekt
    0.41
     і
    0.40
    POSITIVE LOGITS
     реаги
    0.49
     nevoia
    0.48
    请求
    0.48
    কার
    0.45
    询问
    0.44
     চাহি
    0.44
    ğinin
    0.43
     услу
    0.43
     "¿
    0.43
     thử
    0.43
    Act Density 0.003%

    No Known Activations