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

    Neuronpedia

    INDEX
    Explanations

    though

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Basis
    0.91
     คือ
    0.90
     beispielsweise
    0.85
     Interpret
    0.84
     หรือ
    0.82
     yaitu
    0.82
     bijvoorbeeld
    0.79
     Implement
    0.79
     Derivatives
    0.78
     Sebagai
    0.75
    POSITIVE LOGITS
    though
    1.33
    there
    1.28
    featuring
    1.23
    !),
    1.18
    they
    1.18
    whose
    1.15
     featuring
    1.11
     requiring
    1.08
    although
    1.04
    remember
    1.04
    Act Density 0.110%

    No Known Activations