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

    Neuronpedia

    INDEX
    Explanations

    how things behave

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    urdy
    0.41
     போல
    0.41
     Differences
    0.38
     отличие
    0.38
     `<=`
    0.38
     nontrivial
    0.38
    successful
    0.38
    colormap
    0.37
     `<=
    0.36
    urity
    0.36
    POSITIVE LOGITS
     behave
    0.92
     behaves
    0.90
     behaved
    0.88
     behaving
    0.80
    behaved
    0.64
     doing
    0.63
     melakukan
    0.57
    Doing
    0.54
    采取
    0.54
     adopts
    0.52
    Act Density 0.018%

    No Known Activations