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

    Neuronpedia

    INDEX
    Explanations

    safety disclaimers and warnings

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     initial
    0.42
     perceiving
    0.42
     Selection
    0.41
     Selected
    0.40
     INITIAL
    0.40
    Responding
    0.40
     selection
    0.40
    ྷ
    0.40
     Initial
    0.39
     responding
    0.39
    POSITIVE LOGITS
    dangerous
    0.76
     dangereux
    0.75
     dangerous
    0.75
     berbahaya
    0.71
     Dangerous
    0.70
    危险
    0.68
    危険
    0.68
    Dangerous
    0.66
     खतरनाक
    0.65
     peligros
    0.65
    Act Density 0.317%

    No Known Activations