More information can be found on [DBLP, Google Scholar, Semantic Scholar].

Ex uno pluria: Insights on ensembling in low precision number systems Giung Nam, Juho Lee Neural Information Processing Systems (NeurIPS), 2024 (To appear)

Sparse weight averaging with multiple particles for iterative magnitude pruning Moonseok Choi*, Hyungi Lee*, Giung Nam*, Juho Lee International Conference on Learning Representations (ICLR), 2024 [pdf, arXiv]

Enhancing transfer learning with flexible nonparametric posterior sampling Hyungi Lee*, Giung Nam*, Edwin Fong, Juho Lee International Conference on Learning Representations (ICLR), 2024 [pdf, arXiv]

Lipsum-FT: Robust fine-tuning of zero-shot models using random text guidance Giung Nam, Byeongho Heo, Juho Lee International Conference on Learning Representations (ICLR), 2024 [pdf, arXiv]

Traversing between modes in function space for fast ensembling Eunggu Yun*, Hyungi Lee*, Giung Nam*, Juho Lee International Conference on Machine Learning (ICML), 2023 [pdf, arXiv]

Martingale posterior neural processes Hyungi Lee, Eunggu Yun, Giung Nam, Edwin Fong, Juho Lee International Conference on Learning Representations (ICLR), 2023, Spotlight [pdf, arXiv]

Decoupled training for long-tailed classification with stochastic representations Giung Nam*, Sunguk Jang*, Juho Lee International Conference on Learning Representations (ICLR), 2023 [pdf, arXiv]

Benefits of stochastic weight averaging in developing neural network radiation scheme for numerical weather prediction Hwan-Jin Song, Soonyoung Roh, Juho Lee, Giung Nam, Eunggu Yun, Jongmin Yoon, Park Sa Kim Journal of Advances in Modeling Earth Systems (JAMES), October 2022 [pdf, ESSOAr]

Improving ensemble distillation with weight averaging and diversifying perturbation Giung Nam, Hyungi Lee, Byeongho Heo, Juho Lee International Conference on Machine Learning (ICML), 2022 [pdf, arXiv, code]

Diversity matters when learning from ensembles Giung Nam*, Jongmin Yoon*, Yoonho Lee, Juho Lee Neural Information Processing Systems (NeurIPS), 2021 [pdf, arXiv, code]