Email: wuqiyu576 [AT] gmail [DOT] com
Multi-modal NLP and retrieval.
JSPS (The Japan Society for the Promotion of Science) Research Fellowship for Young Scientists (DC2). This fellowship program is Japan’s core program for cultivating young researchers.
Reviewer of ACL Rolling Review (2023, 2024); ICLR (2022, 2023, 2024); EMNLP 2023; NeurIPS (2021, 2022, 2023); ICML 2022.
During this internship at Creative AI Lab at Sony, I extend my research scope to NLP for cross-modal representation learning. I proposed BiAug to mitigate reporting bias problem in visual-language datasets with bimodal augmentation.
I proposed PCL at EMNLP 2022 that tackles the bias problem in the sentence embeddings during the internship. The codes and models can be found at https://github.com/qiyuw/PeerCL
I proposed TNF at ICLR 2021 to mitigate the rare word problem in the language pre-training. We use an external memory to store and provide richer information for rare words on the fly.
I have contributed to the transportation recommendation systems of Baidu Maps, and worked on community-aware transportation prediction AAAI 2021 (co-first author) during the internship.
JST (Japan Science and Technology Agency) Support for Pioneering Research Initiated by the Next Generation (SPRING) Program, The University of Tokyo “Advanced Human Resource Development Leading Green Transformation (GX) (SPRING GX)” Project Student.
TA of Social Implications of Computing at Peking University.