MACa: Mathematics in Generative AI
Mon, 28 Oct, 08:15 - 09:55 PT (UTC -8)
Location: Chapel
Session Type: Lecture
Session Chair: Qing Qu, University of Michigan
Track: Adaptive Systems, Machine Learning, and Data Analytics
Mon, 28 Oct, 08:15 - 08:40 PT (UTC -8)

MACa.1: White-Box Transformers via Sparse Rate Reduction
Yaodong Yu, University of California, Berkeley, United States; Sam Buchanan, Toyota Technological Institute at Chicago, United States; Druv Pai, University of California, Berkeley, United States; Tianzhe Chu, ShanghaiTech University, China; Ziyang Wu, University of California, Berkeley, United States; Shengbang Tong, New York University, United States; Ben Haeffele, Johns Hopkins University, United States; Yi Ma, University of California, Berkeley and University of Hong Kong, Hong Kong SAR of China
Mon, 28 Oct, 08:40 - 09:05 PT (UTC -8)

MACa.2: Seeking the principles of unsupervised representation learning
Yubei Chen, University of California, Davis, United States
Mon, 28 Oct, 09:05 - 09:30 PT (UTC -8)

MACa.3: Self-Play Preference Optimization for Language Model Alignment
Yue Wu, UCLA, United States; Zhiqing Sun, CMU, United States; Huizhuo Yuan, Kaixuan Ji, UCLA, United States; Yiming Yang, CMU, United States; Quanquan Gu, UCLA, United States
Mon, 28 Oct, 09:30 - 09:55 PT (UTC -8)

MACa.4: Understanding Diffusion Models via Low-Dimensional Modeling
Peng Wang, Huijie Zhang, Zekai Zhang, Yi Ma, Qing Qu, University of Michigan, United States