I’m obsessed with exciting problems on knowledge mining from unstructured data with real applications in real scenarios (e.g., with limited resource and human annotation). I worked on the information extraction module of the Aminer system at Tsinghua University (ASONAM’16, SNAM’18). I spent three summers at Google, building practical tools and systems on relation extraction, news story headline generation (WWW’20), and accelerating large-scale language model pre-training (NAACL’21).