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

    Neuronpedia

    INDEX
    Explanations

    additional gradients calculation

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Antique
    0.82
     antique
    0.75
     erfolgre
    0.75
     مو
    0.74
     Doors
    0.74
     speeds
    0.73
     guessed
    0.73
     doors
    0.72
     GPS
    0.70
     sketchy
    0.70
    POSITIVE LOGITS
     measures
    1.56
    衡量
    1.52
     metrics
    1.49
    measures
    1.46
     measure
    1.45
     Measures
    1.39
     calculer
    1.39
     Metrics
    1.38
    Calculation
    1.37
     calcular
    1.36
    Act Density 0.709%

    No Known Activations