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

    Neuronpedia

    INDEX
    Explanations

    politeness and related phrases

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    tle
    1.16
    tida
    1.05
    MO
    1.03
    ines
    1.02
    ilization
    1.00
    它
    1.00
    昉
    0.99
    oys
    0.99
    ibur
    0.99
    iya
    0.98
    POSITIVE LOGITS
     sloppy
    1.10
    িন
    1.02
    ور
    1.02
     tranche
    1.01
     hurry
    0.98
     lousy
    0.93
    issance
    0.91
     defenseman
    0.89
     mouthful
    0.89
     a
    0.88
    Act Density 0.028%

    No Known Activations