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

    Neuronpedia

    INDEX
    Explanations

    doesn't just say

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     three
    1.71
     four
    1.70
     two
    1.68
     (
    1.60
     new
    1.57
     short
    1.54
    1.53
     *
    1.51
     same
    1.48
     still
    1.48
    POSITIVE LOGITS
    okatokat
    3.25
    眎
    3.15
    ꗟ
    3.13
     Pusenkoff
    3.11
    𒆗
    3.08
    𒅇
    3.05
    毉
    3.05
    𒐊
    3.05
     trataro
    3.04
    ഫിഈ
    3.04
    Act Density 0.430%

    No Known Activations