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

    Neuronpedia

    INDEX
    Explanations
    No Explanations Found
    New Auto-Interp
    Negative Logits
     Whenever
    1.09
    Whenever
    1.04
     sogenannten
    0.90
     sogen
    0.86
     sogenannte
    0.85
    whenever
    0.85
    <unused2197>
    0.84
    することができる
    0.84
    <unused2165>
    0.84
    Todas
    0.84
    POSITIVE LOGITS
    ..
    1.34
    ↵↵↵
    1.27
    ↵↵↵↵
    1.21
    ↵
    1.21
    !.
    1.15
    .}
    1.14
     etc
    1.11
    .(
    1.10
    ↵↵
    1.08
     :)
    1.07
    Act Density 1.983%

    No Known Activations