杨轶,美国加州大学伯克利分校博士后。本科就读于北京师范大学天文系，毕业后前往美国得克萨斯农工大学继续求学，2017年获得物理学博士学位。2017年至2020年在以色列Weizmann科学院做博士后。现为美国加州大学伯克利分校博士后研究员。他的主要研究方向包括超新星的观测，大视场巡天，以及星际介质。The explosions of supernovae are so powerful that the ejecta wipe out almost all traces of the pre-explosion configuration within very few days. Conventional photometry or spectroscopy provides limited information on the structures of the circumstellar matter as well as the SN ejecta themselves. This is because such information is projected and smeared into the single dimension of radial velocity. By contrast, critical information about the interaction of the ejecta with any companion and/or any circumstellar matter is significantly encoded in the early polarization spectra. Thanks to recent enormous advances in supernova survey strategies, we can probe the geometric properties of the Type Ia SNe explosion at unprecedentedly early phases. I will also discuss the lessons we would expect to learn from more massive core-collapse supernovae at extremely early stages.