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

    Neuronpedia

    INDEX
    Explanations

    comparisons and degrees of difference

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     diferentes
    0.50
     irgendwie
    0.48
    ichever
    0.48
     Lots
    0.46
    得很
    0.46
    Increasing
    0.45
     different
    0.44
    ほとんど
    0.44
     somehow
    0.44
     różnych
    0.43
    POSITIVE LOGITS
     quite
    0.85
    Quite
    0.77
    quite
    0.75
     Quite
    0.73
     comparable
    0.70
     compar
    0.66
     anywhere
    0.64
     igual
    0.63
     equivalently
    0.61
     столь
    0.57
    Act Density 0.027%

    No Known Activations