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

    Neuronpedia

    INDEX
    Explanations

    nonsensical text

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
    valu
    -0.07
    trained
    -0.07
     mess
    -0.06
     �
    -0.06
    meyen
    -0.06
     Alberta
    -0.06
     плеч
    -0.06
     harvest
    -0.06
    zas
    -0.06
    -content
    -0.06
    POSITIVE LOGITS
    ,");↵
    0.07
    Disappear
    0.06
    uppy
    0.06
    ++
    0.06
     CSR
    0.06
    __:
    0.06
    Large
    0.06
    HAM
    0.06
    ância
    0.06
    ]]↵↵
    0.06
    Act Density 0.042%

    No Known Activations