Aaron Wang Email & Phone Number

New York, Ny, Us

New York, Ny, Us

Kirkland, Wa, Us

Batavia, Il, Us

• Fully funded graduate fellowship in partnership with FNAL and ANL for computational particle physics research• Researching efficient transformer models for the L1 trigger at the CERN CMS experiment• Using deep learning to develop efficient anomaly detection and jet tagging software• Managing O2O system for the CMS experiment, doing MC GT updates and facilitating smooth experiment runs

Chicago, Il, Us

• Used deep learning to predict secondary structures on membrane binding proteins• Curated dataset of membrane binding amphipathic helix proteins

Chicago, Il, Us

• Developing GAN models to generate novel superconductive materials• Developing a multi-objective optimization tool for creation of new biocides using parallel multi-objectiveoptimization (ParMOO)• Studying data science based approaches for topological materials

Seattle, Wa, Us

• Read machine learning literature, and then adapted novel machine learning models for particle physics applications• Deployed deep learning models in PyTorch and TensorFlow in order to research and optimize 5-class particle jet taggers• Developed recursive and recurrent machine learning models for adaptation into HLS4ML• Developed specialized transformer model for use in sequential particle jet tagging• Gave bi-weekly research talks to graduate students and the Principle Investigator

Seattle, Wa, Us

• Grew various quantum material crystals using flux, vapor, and Bridgman growth methods under Professor Jiun-Haw Chu• Prepared samples of various materials, and took resistivity versus temperature measurements on the materials using the Quantum Design DynaCool PPMS system• Performed analysis on resistivity data in order to identify quantum critical points and novel material properties• Collaborated with graduate students to design and improve growth recipes for novel materials

Richland, Wa, Us

• Developed and optimized few-shot models for materials solubility prediction using Pytorch• Conducted statistical analysis on model performance in order to determine molecule types that are the most difficult to predict, and discover underlying physics/chemistry• Used spearman correlation in order to determine the importance of feature information, reducing the number of features used by 30 percent while maintaining model performance.

Princeton, Nj, Us

IRIS-HEP Fellow• Developed software for use in high energy physics as part of the IRIS-HEP program• Currently working on making the ROB(Reproducible Open Benchmarks) compatible with python notebooks• Used python libraries such as papermill to package jupyter notebooks so that the ROB can be run on jupyter notebooks• Developed demos to demonstrate the usage of the ROB within particle physics and machine learning• Presented at the ML4Jets and APS DPF Conference

Seattle, Wa, Us

•Received $3000 dollars in grant money for nanofabrication from the Washington Research Foundation•Used clean room devices to prototype, then manufacture cantilevers for use in material characterization•Conducted research on the measurement of structural transitions in quantum materials using cantilevers

Doctor Of Philosophy - Phd, Physics

Bachelor Of Science - Bs, Comprehensive Physics

High School Diploma