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

    Neuronpedia

    INDEX
    Explanations

    kena, suay, hacked, getroffen

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Hearst
    0.38
     unhealthy
    0.37
     emergent
    0.37
    عبير
    0.37
    чества
    0.36
     শির
    0.36
    ịnh
    0.36
    честве
    0.35
    ൩
    0.35
    다운
    0.34
    POSITIVE LOGITS
     terken
    0.69
     kena
    0.61
     getroffen
    0.56
     pata
    0.55
    hab
    0.54
     robbed
    0.50
     puk
    0.49
     trampled
    0.49
     hab
    0.48
     geb
    0.48
    Act Density 0.001%

    No Known Activations