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

    Neuronpedia

    INDEX
    Explanations

    steep drops and edges

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
     sensors
    -0.07
     shields
    -0.06
    )$
    -0.06
     adultery
    -0.06
     heavens
    -0.06
     interpreted
    -0.06
     scenario
    -0.06
     tract
    -0.06
     близько
    -0.06
    pawn
    -0.06
    POSITIVE LOGITS
    aine
    0.07
     blamed
    0.07
    ází
    0.06
     ظرفیت
    0.06
    ube
    0.06
    -padding
    0.06
     mediator
    0.06
    生命
    0.06
    \xff
    0.06
     LCS
    0.06
    Act Density 0.013%

    No Known Activations