On March 22, 2023, two graduate students from the Tang Chao Laboratory participated in the American Physical Society (APS) March Meeting held in Las Vegas from March 5th to March 10th. On the morning of March 7th, Yang Ning gave a presentation entitled "Stochastic gradient descent introduces an effective landscape-dependent regularization favoring flat solutions", which mainly explained how stochastic gradient descent (SGD) helps finding flatter (more generalizable) solutions in the loss landscape of overparameterized neural networks from the perspective of statistical physics. For more information, please refer to: https://meetings.aps.org/Meeting/MAR23/Session/F02.4.
Founded in 1899, the American Physical Society (APS) is one of the largest physics organizations in the world, dedicated to promoting the development and application of physics. This conference covers multiple fields of physics, including but not limited to particle physics, condensed matter physics, biophysics, astrophysics, nuclear physics, and atomic physics. We mainly participated in the conferences related to biophysics and machine learning, and the related conference materials can be found at: https://disk.pku.edu.cn:443/link/AB0F247AEA348543077C93CB9B54A064.
