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

    Neuronpedia

    INDEX
    Explanations

    code delimiters and keywords

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     entertaining
    0.34
     comforting
    0.33
     sleep
    0.33
     sofa
    0.32
     bertanggung
    0.31
     comforted
    0.31
     Angel
    0.31
     ada
    0.31
     firefox
    0.31
     bartender
    0.31
    POSITIVE LOGITS
    сю
    0.34
    tional
    0.33
    ре
    0.32
    ност
    0.32
    ture
    0.31
    реи
    0.31
     issus
    0.31
    ﷽
    0.30
    yzed
    0.30
     dey
    0.30
    Act Density 2.129%

    No Known Activations