Hi everyone, We’d like to invite you to submit to the NeurIPS 2023 Workshop on Distribution Shifts: New Frontiers with Foundation Models.
Website: https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fsites.google.com%2Fview%2Fdistshift2023&data=05%7C01%7Cuai%40ENGR.orst.edu%7C3a5b8581dc354c2173c208dbae72e61c%7Cce6d05e13c5e4d6287a84c4a2713c113%7C0%7C0%7C638295581322844463%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C2000%7C%7C%7C&sdata=RlfP8rZLpMmUcIpeWfw9lmXbv%2BnzTXXwoy9mPnLxvic%3D&reserved=0 Paper submission deadline: October 2, 2023 (Anywhere on Earth) Author notification: October 27, 2023 Workshop: December 15, 2023, in-person in New Orleans, USA. Authors who will not be able to attend in person are still encouraged to submit. Accepted papers will be accompanied by a short pre-recorded video to allow authors to present their work remotely. Please reach out to distshift-workshop-2...@googlegroups.com if you have any questions. *Call for papers* Distribution shifts—where a model is deployed on a data distribution different from what it was trained on—pose significant robustness challenges in real-world ML applications. Such shifts are often unavoidable in the wild and have been shown to substantially degrade model performance. For example, models can systematically fail when tested on patients from different hospitals or people from different demographics. Training models that are robust to such distribution shifts is a rapidly growing area of interest in the ML community. In recent years, foundation models—large pretrained models that can be adapted for a wide range of tasks—have achieved unprecedented performance on a broad variety of discriminative and generative tasks, including in distribution shift scenarios. Foundation models open up an exciting new frontier in the study of distribution shifts.The goal of our workshop is to foster discussions and further research on distribution shifts, especially in the context of foundation models. Examples of relevant topics include, but are not limited to: Effects of foundation models (e.g., pre-training, scale) on robustness Robust adaptation of foundation models to downstream tasks Distribution shifts from pretraining to downstream distributions, including in the context of generative foundation models. Beyond the above topics, we are broadly interested in methods, empirical studies, and theory of distribution shifts, including those that do not involve foundation models. *Invited speakers* Aditi Raghunathan, Carnegie Mellon University Balaji Lakshminarayanan, Google DeepMind Hoifung Poon, Microsoft Research Kate Saenko, Boston University Ludwig Schmidt, University of Washington Peng Cui, Tsinghua University *Organizers* Becca Roelofs, Google Fanny Yang, ETH Zurich Hongseok Namkoong, Columbia University Jacob Eisenstein, Google Masashi Sugiyama, RIKEN & University of Tokyo Pang Wei Koh, University of Washington Shiori Sagawa, Stanford University Tatsunori Hashimoto, Stanford University Yoonho Lee, Stanford University
_______________________________________________ uai mailing list uai@engr.orst.edu https://it.engineering.oregonstate.edu/mailman/listinfo/uai
