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

    Neuronpedia

    INDEX
    Explanations

    How things work

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
     HOW
    -0.07
    Top
    -0.06
     helps
    -0.06
     Opens
    -0.06
    .lr
    -0.06
     університ
    -0.06
    .Relative
    -0.06
     tat
    -0.06
     When
    -0.06
    ко
    -0.06
    POSITIVE LOGITS
     być
    0.07
    _regions
    0.07
    asions
    0.07
    backward
    0.06
    Ult
    0.06
     Entity
    0.06
     боль
    0.06
     Phantom
    0.06
     prost
    0.06
    *size
    0.06
    Act Density 0.184%

    No Known Activations