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

    Neuronpedia

    INDEX
    Explanations

    this and that references

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     As
    0.42
     ಹೆ
    0.38
     sometime
    0.38
    As
    0.37
    iziModal
    0.37
     Unfortunately
    0.35
    ger
    0.35
    有一种
    0.35
    ilogue
    0.35
     easier
    0.35
    POSITIVE LOGITS
     THIS
    1.23
    this
    1.09
     this
    1.07
    THIS
    1.04
     هذا
    1.03
     THAT
    1.02
     이렇게
    1.00
     этого
    0.98
    이렇게
    0.96
     цього
    0.93
    Act Density 0.024%

    No Known Activations