Posted 1/15/2024, 4:10:05 PM
Machine learning model reveals and compensates for hidden imperfections in quantum materials
- Research team develops ML model to address variability in quantum devices caused by material imperfections
- Model analyzes electron flow to infer internal disorder patterns, improving performance predictions
- Approach combines physics simulations and deep learning to quantify and compensate for device variability
- Allows more accurate engineering of quantum materials by revealing previously invisible nanoscale disorders
- Demonstrates utility of physics-aware ML to bridge "reality gap" between predicted and observed device behaviors