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

    Neuronpedia

    INDEX
    Explanations

    safety guidelines

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    安全
    0.50
    安全性
    0.47
    Safety
    0.46
     안전
    0.45
     безпе
    0.45
    safety
    0.44
     安全
    0.44
    SAFETY
    0.44
     Sicherheit
    0.43
     безопасности
    0.43
    POSITIVE LOGITS
     guidelines
    0.49
     protocols
    0.47
     barriers
    0.47
     protocolos
    0.45
    protocols
    0.41
     Barriers
    0.39
     Protocols
    0.38
     Guidelines
    0.38
    ető
    0.38
    одо
    0.38
    Act Density 0.029%

    No Known Activations