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

    Neuronpedia

    INDEX
    Explanations

    explicitly prohibited

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Safety
    0.50
     safety
    0.46
    安全
    0.46
     parfois
    0.46
    ethics
    0.46
    Boundary
    0.46
     ethics
    0.45
     verantwort
    0.44
     सुरक्षित
    0.44
    please
    0.43
    POSITIVE LOGITS
     Explicit
    0.55
     Detailed
    0.48
     goes
    0.48
     explicitly
    0.47
     contradicts
    0.47
     explicit
    0.47
     Goes
    0.46
     Platforms
    0.46
     LI
    0.45
     Directly
    0.45
    Act Density 0.008%

    No Known Activations