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

    Neuronpedia

    INDEX
    Explanations

    Market order

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     dritten
    0.30
     problème
    0.28
     ربعه
    0.28
    发动
    0.27
     affirme
    0.27
     terrib
    0.27
     едва
    0.27
    놁
    0.27
     selben
    0.27
    ‹
    0.26
    POSITIVE LOGITS
     and
    0.49
    and
    0.40
    5
    0.37
    8
    0.36
    _
    0.35
     or
    0.35
    1
    0.35
    /
    0.34
    4
    0.33
    7
    0.33
    Act Density 0.381%

    No Known Activations