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

    Neuronpedia

    INDEX
    Explanations

    next important step

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     initially
    0.75
    setDefault
    0.74
     Initially
    0.70
    Initially
    0.69
    unan
    0.67
    当初
    0.65
    lesssim
    0.64
    Originally
    0.64
     originally
    0.63
    laus
    0.61
    POSITIVE LOGITS
     step
    3.41
     steps
    3.04
    step
    2.99
     Step
    2.95
    Step
    2.87
     Steps
    2.77
    steps
    2.59
    ステップ
    2.58
    Steps
    2.55
     STEP
    2.48
    Act Density 0.398%

    No Known Activations