Chun-Teh Chen

Research background: My research journey began at National Taiwan University (NTU), where I pursued a master’s degree under the guidance of Prof. Yeong-Bin Yang, focusing on the vibration analysis of a symmetric two-member truss. While a relatively simple governing equation could describe this system, it exhibited chaotic behavior when accounting for geometric nonlinearity. This phenomenon fascinated me, illustrating how complexity can emerge from simplicity, often called the butterfly effect. This initial experience with the intricate behavior of deterministic systems sparked my motivation to pursue a Ph.D. in Structures and Materials at MIT. At NTU, my research involved the finite element method (FEM) and structural dynamics, but at MIT, I shifted my focus toward atomistic modeling and quantum chemical calculations. Working with Prof. Markus Buehler, I explored the structural, mechanical, and optical properties of natural materials like eumelanin (a pigment found in many organisms) and polydopamine (a mussel-inspired material for multifunctional coatings) through multiscale modeling. This technique integrates different simulation methods to understand material behavior across multiple scales—linking atomic-scale phenomena to macro-scale properties. At MIT, I gained expertise in atomistic modeling, particularly molecular dynamics (MD) simulations and density functional theory (DFT). My research findings were published in top-tier journals such as ACS Nano, Nature Communications, and Chemical Science, and were featured in MIT News.

​

Passion for artificial intelligence: In 2016, I was inspired by the groundbreaking success of Google DeepMind’s AlphaGo, which highlighted the immense potential of artificial intelligence (AI) and machine learning (ML) in advancing scientific computing. During my postdoctoral training at MIT, I collaborated with Prof. Markus Buehler on two pioneering papers on data-driven composite design. These papers, published in Extreme Mechanics Letters and Materials Horizons in 2018, have each garnered over 400 citations (as of August 2024), with the latter receiving the Materials Horizons Outstanding Paper Award. I continued to build on my work in ML applications at UC Berkeley, publishing five additional papers between 2019 and 2023. In 2019, we developed a data-driven approach for designing high-performance graphene nanocomposites, featured on the cover of Advanced Theory and Simulations. That same year, we published a forward-looking paper on the role of ML in composite materials design in MRS Communications, which earned the MRS Communications Lecture Award. In 2020, we introduced a versatile inverse design framework, generative inverse design networks (GIDNs), published in Advanced Science. This framework has since inspired numerous follow-up studies across various disciplines and has accrued over 260 citations (as of August 2024). In 2021, we launched ElastNet, an innovative elastography method that integrates elasticity theory with deep learning. Published in PNAS, this work demonstrated that representing physical quantities as continuous, differentiable functions via neural networks offers significant advantages over conventional grid-based techniques. Most recently, in 2023, we expanded ElastNet’s capabilities to address complex forward, inverse, and mixed problems, with these advancements published in Advanced Science.

Acknowledgments: I am deeply grateful for the mentorship from my advisors — Prof. Yeong-Bin Yang at NTU, Prof. Markus Buehler at MIT, and Prof. Daryl Chrzan and Prof. Mark Asta at UC Berkeley. Additionally, I appreciate the financial support that enabled my studies and research. During my Ph.D. at MIT, I received a fellowship from the Taiwanese Government and funding from CRP Henri Tudor for the BioNanotechnology project. My postdoctoral research at MIT was supported by the Office of Naval Research (ONR) and the Multidisciplinary University Research Initiative (MURI). My postdoctoral research at UC Berkeley was funded by the College of Engineering and the Department of Mechanical Engineering at UC Berkeley.