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

    Neuronpedia

    INDEX
    Explanations

    versions, variations, methods, vulnerabilities, strains

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    하지만
    0.50
    :
    0.48
    .
    0.45
     a
    0.42
     mutta
    0.41
     maar
    0.40
    だけど
    0.39
     an
    0.39
     hết
    0.38
    ?
    0.38
    POSITIVE LOGITS
     were
    0.81
     occurs
    0.73
     are
    0.72
     is
    0.70
     was
    0.67
     occurred
    0.67
     occur
    0.65
     emerges
    0.64
     may
    0.64
     would
    0.62
    Act Density 0.021%

    No Known Activations