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

    Neuronpedia

    INDEX
    Explanations

    Bypassing defenses

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
     نار
    -0.07
     pyplot
    -0.07
     filles
    -0.06
     Nixon
    -0.06
     конце
    -0.06
     dbc
    -0.06
     xor
    -0.06
    社
    -0.06
     alternate
    -0.06
    -runner
    -0.06
    POSITIVE LOGITS
    �
    0.07
    的地
    0.06
    .DropDown
    0.06
    осред
    0.06
     handwriting
    0.06
    .edit
    0.06
    ！");↵
    0.06
     slun
    0.06
    جمع
    0.06
     makes
    0.06
    Act Density 0.027%

    No Known Activations