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

    Neuronpedia

    INDEX
    Explanations

    improving text or prompts

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    Dynamics
    0.40
    dynamics
    0.39
     dynamics
    0.38
    ချိန်
    0.35
     Dynamics
    0.35
     unreported
    0.34
     Therapeutics
    0.34
    ವಿಧ
    0.34
     erlebt
    0.33
    ≙
    0.33
    POSITIVE LOGITS
     улуч
    0.70
     improve
    0.66
    完善
    0.65
     개선
    0.64
     Improve
    0.63
     migliorare
    0.63
     améliorer
    0.62
    améliorer
    0.61
     улучшения
    0.61
    improve
    0.60
    Act Density 0.007%

    No Known Activations