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

    Neuronpedia

    INDEX
    Explanations

    consistently ranked

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     perhaps
    0.94
     maybe
    0.90
     pretending
    0.87
     adventures
    0.85
     hopefully
    0.83
     trying
    0.83
     misschien
    0.82
    maybe
    0.81
     really
    0.80
     specialise
    0.79
    POSITIVE LOGITS
     Ranked
    1.21
     ranked
    1.20
    ranks
    1.11
    Recognition
    1.11
     rankings
    1.09
    ranked
    1.08
    recognized
    1.08
     Recognition
    1.08
     rank
    1.08
     ranks
    1.08
    Act Density 0.114%

    No Known Activations