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

    Neuronpedia

    INDEX
    Explanations

    AI capabilities and actions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    삭제
    0.79
     quantifier
    0.71
     SINGLE
    0.66
     nondegenerate
    0.65
     삭제
    0.64
     cannot
    0.64
     incapable
    0.64
     universality
    0.64
     සී
    0.63
     Unable
    0.62
    POSITIVE LOGITS
     engage
    1.01
     communicate
    0.88
     Engage
    0.86
    communicate
    0.85
    oodle
    0.82
     carry
    0.81
    engage
    0.81
    Manip
    0.77
     convey
    0.76
     manipulate
    0.75
    Act Density 0.388%

    No Known Activations