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

    Neuronpedia

    INDEX
    Explanations

    really focus on savoring

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     enjoy
    0.72
     enjoying
    0.70
    enjoy
    0.60
    Enjoy
    0.57
    楽し
    0.55
     enjoys
    0.55
     enjoyed
    0.54
     enjoyment
    0.52
     disfrutar
    0.52
    楽しく
    0.51
    POSITIVE LOGITS
     fully
    0.99
     Fully
    0.98
    Fully
    0.95
     truly
    0.82
     Really
    0.74
     immersing
    0.74
     Truly
    0.73
    Truly
    0.72
     Properly
    0.71
     properly
    0.70
    Act Density 0.021%

    No Known Activations