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

    Neuronpedia

    INDEX
    Explanations

    piecewise functions and intervals

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
    tow
    -0.10
    Tow
    -0.08
     Diplom
    -0.08
     lag
    -0.08
     thrift
    -0.08
     seen
    -0.07
    rena
    -0.07
    Dipl
    -0.07
     ade
    -0.07
     diplomacy
    -0.07
    POSITIVE LOGITS
     thresholds
    0.11
    处分
    0.11
     milestones
    0.10
     divides
    0.09
     partitions
    0.09
     jun
    0.09
     segmentation
    0.08
    划
    0.08
    Threshold
    0.08
     Threshold
    0.08
    Act Density 0.026%

    No Known Activations