Yangyang Li

• Led simulation efforts for developing modular physics-driven SEM emulator, driving system-level emulation into digital twins of SEM. • Built GPU-accelerated particle tracing software in C++/CUDA for electron optics design, achieving 137× speedup over legacy tools. • Enhanced reticle inspection algorithms using high-fidelity simulation, improving defect detection accuracy and boosting synthetic database generation for SEM-based platforms. • Partnered with cross-functional teams to validate new hardware features via emulation, cutting POR turnaround time and accelerating feature trials. • Engaged with marketing to handle customer request on the technical path finding of inspection for new high-NA(high-k/low-n) reticle materials

• Implemented customized computational graph with TensorFlow for critical dimension small-angle X-ray scattering (CD-SAXS) metrology on high-aspect-ratio 3D NAND/DRAM. • Completed geometry discretization of complex 3D NAND and computation of Fourier transformation. • Completed sensitivity analysis of the variation of the signal induced by physical parameters. • Coordinate among team members for requirement analysis, feasibility analysis, tool development, and feature request. • Perform robust unit test, integration test, and validating test with simulated signal from physics first principle

• Applied stationary iterative methods to simulate stellar spectra under the assumption of non-local thermodynamical equilibrium. • Developed open-source code to predict features from observational data, with detailed documentation of data ETL, visualization, and API framework. • Build CNN to emulate stellar spectra with multiple labels with great accuracy

Fall 2018: AST1002 - Discovering the Universe AST1022L - Astronomy Laboratory Spring 2019: AST3043 - History of Astronomy through Newton Fall 2019: AST2037 - Life in the Universe Spring 2020: AST4300 - Galactic Astronomy

• Applied Gaussian Process (GP) to analyze time series of stars from Kepler telescope • Implemented statistic methods infer the distribution of physics parameters from GP • Optimized the run time of the pipeline with great improvement in speed • Trained a Convolutional Neural Network (CNN) to classify exoplanets with great precision • Identified several exoplanet candidates in the Kepler data sample and presented results

The Summer Undergraduate Research Fellowships (SURF) program

I was a member of the School of the Gifted Young.