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

    Neuronpedia

    INDEX
    Explanations

    safety and security considerations

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    toire
    0.65
     wording
    0.65
     Relevant
    0.64
    ૈય
    0.64
    Коммента
    0.62
    MVCProject
    0.62
    chsler
    0.62
    ਦਰ
    0.62
     बोस
    0.61
    禅
    0.61
    POSITIVE LOGITS
     safe
    4.29
     safely
    3.92
    safe
    3.84
     Safe
    3.80
    Safe
    3.76
    安全
    3.71
     безопас
    3.56
     safety
    3.46
     안전
    3.44
     safer
    3.43
    Act Density 0.654%

    No Known Activations