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

    Neuronpedia

    INDEX
    Explanations

    Code arguments

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
     placing
    -0.07
    奔
    -0.07
     raising
    -0.07
     copying
    -0.07
     يول
    -0.07
     from
    -0.07
    �
    -0.07
     stehen
    -0.07
     réc
    -0.07
     hortic
    -0.07
    POSITIVE LOGITS
     CCS
    0.09
     Geen
    0.09
    particip
    0.09
    �襪
    0.08
    ১
    0.08
    �ย
    0.08
    ুদ্র
    0.08
    �다
    0.08
     XK
    0.08
    👉
    0.08
    Act Density 0.001%

    No Known Activations