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

    Neuronpedia

    INDEX
    Explanations

    account for deviations

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    CollectionView
    0.40
    财经
    0.39
     collectionView
    0.39
    轅
    0.38
     auditions
    0.38
     paycheck
    0.38
    tLogRow
    0.38
     dedication
    0.38
     payroll
    0.37
     playbook
    0.37
    POSITIVE LOGITS
     nonlinearity
    0.86
     inhomogeneities
    0.86
     noise
    0.81
     heterogeneity
    0.79
     effects
    0.77
     inaccuracies
    0.76
     distortions
    0.75
     variability
    0.75
     fluctuations
    0.75
     imperfections
    0.74
    Act Density 0.074%

    No Known Activations