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

    Neuronpedia

    INDEX
    Explanations

    links and link following common words

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    𝑢
    -1.06
     with
    -1.02
    也不
    -0.99
    sə
    -0.99
    gale
    -0.98
    水墨
    -0.98
     chão
    -0.98
     escritório
    -0.95
    "}";
    -0.95
    -0.94
    POSITIVE LOGITS
     to
    1.52
     link
    1.27
     from
    1.20
     into
    1.17
     links
    1.16
     menuju
    1.13
     on
    1.07
    考核
    1.07
    href
    1.07
    ḗ
    1.03
    Act Density 0.040%

    No Known Activations