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

    Neuronpedia

    INDEX
    Explanations

    using glue as adhesive

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    d
    -2.64
     meeste
    -2.64
    ᎑
    -2.56
    泺
    -2.48
    t
    -2.45
    </i>
    -2.42
     Это
    -2.31
     egregious
    -2.30
     meisten
    -2.30
     работа
    -2.30
    POSITIVE LOGITS
    鷥
    2.41
     serão
    2.27
     verlangen
    2.25
     marinho
    2.25
     gewi
    2.20
     gewor
    2.11
    眎
    2.06
    圌
    2.05
    胧
    2.05
     Pág
    2.05
    Act Density 0.029%

    No Known Activations