Yang Weifeng, born in 1977 in Zibo, Shandong Province, is a professor at the School of Physics and Optoelectronic Engineering and the Center for Theoretical Physics at Hainan University. His research primarily focuses on strong-field atomic and molecular physics and attosecond science. A major achievement of his research is the development, together with his team, of the deep-learning Feynman path integral method for strong-field dynamics. It has overcome key technical bottlenecks in ultrafast spatiotemporal dynamics — including the enormous volume of analytical computation data and low information resolution — and pioneered a new direction at the intersection of machine learning and attosecond physics.
In 2004, Yang Weifeng was admitted to the Shanghai Institute of Optics and Fine Mechanics, Chinese Academy of Sciences (CAS), to pursue his doctoral studies under the supervision of Academician Xu Zhizhan and Professor Gong Shangqing, both leading experts in China’s strong-field ultrafast physics. The early stages of his research proved quite challenging, but he never backed down. “For the things I’m passionate about, my secret is an inexhaustible enthusiasm,” he said. He devoted every available moment to his work, and years of prolonged sitting at his desk led to a severe lumbar disc herniation. Today, he props his computer up high and works standing at his desk. Before graduation, his findings on the carrier-envelope phase effect during the propagation of cycle-scale laser pulses in polarized media were published in a prestigious physics journal, and his doctoral dissertation was awarded the National Excellent Doctoral Dissertation. In 2011, he went abroad for postdoctoral research, where he continued to explore new frontiers.
Attosecond science seeks to capture the ultrafast motion of electrons (one attosecond is one quintillionth of a second) in the microscopic world. Traditional exhaustive path algorithms require calculating every individual trajectory of electron motion, resulting in enormous data volumes and computational demands. One day, quite by chance, Yang Weifeng drew inspiration from a sci-fi movie scene depicting a “clone rehearsing every possible outcome,” which sparked the idea of using artificial neural networks to achieve precise predictions of exhaustive results. Around 2016, he and his team seized the opportunity presented by the rapid iteration of artificial intelligence. After hundreds of thousands of rounds of verification, failure, and re-verification, they achieved a major breakthrough in 2020. They developed for the first time internationally, a deep-learning-based strong-field dynamics analysis method built on Feynman path integrals, attaining a speedup of more than two orders of magnitude and pioneering a new direction at the deep intersection of machine learning and attosecond physics. For example, processing 10 billion electron trajectory data points requires at least two months on the most advanced supercomputer using the traditional exhaustive method, whereas the new approach can complete the same task in just a few hours. A team of international scientists commented that this achievement “has resolved the formidable challenge of exhaustively sampling all paths in electron dynamics.”
Yang Weifeng has a deep passion for teaching. He typically prepares his lectures two to three weeks in advance, weaving scenes from books such as The Three-Body Problem and Harry Potter, along with current academic hot topics, into his classroom sessions. “I hope my students will always stay curious and experience the beauty of physics,” he said. Rigorous in scholarship and devoted to research, he has mentored outstanding students. His first doctoral graduate, Liu Xiwang, is now an associate professor at the School of Physics and Optoelectronic Engineering at Hainan University. In 2023, Yang underwent heart surgery and spent three days in the intensive care unit. Less than a month after the operation, he traveled to Yinchuan, capital city of Ningxia Hui Autonomous Region in northwestern China, to attend and deliver a talk at the Fall Meeting of the Chinese Physical Society, where he successfully secured the hosting rights for Hainan University to hold the 2024 Fall Meeting of the Chinese Physical Society.
Translated by Tang Zijie
Proofread by Chen Chuanxian
