I am a PhD student in KAIST School of Computing, supervised by Professor Seunghoon Hong. My current research interest is few-shot learning and general-purpose AI. In particular, I am interested in building a generalist model to solve various tasks without heavy task-specific training.

Recent Publications

* denotes equal contribution.

[1] Donggyun Kim, Chanwoo Kim, Seunghoon Hong, "HyperFlow: Gradient-Free Emulation of Few-Shot Fine-Tuning", arXiv preprint, 2025. [paper]

[2] Chanhyuk Lee, Jiho Choi, Chanryeol Lee, Donggyun Kim, Seunghoon Hong, "AdaRank: Adaptive Rank Pruning for Enhanced Model Merging", arXiv preprint, 2025. [paper]

[3] Jiho Choi, Donggyun Kim*, Chanhyuk Lee, Seunghoon Hong, "Revisiting weight averaging for model merging", arXiv preprint, 2024. [paper]

[4] Seongwoong Cho*, Donggyun Kim*, Jinwoo Lee, Seunghoon Hong, "Meta-Controller: Few-Shot Imitation of Unseen Embodiments and Tasks in Continuous Control", Advances in neural information processing systems (NeurIPS), 2024. [paper] [github]

[5] Donggyun Kim, Seongwoong Cho, Semin Kim, Chong Luo, Seunghoon Hong, "Chameleon: A Data-Efficient Generalist for Dense Visual Prediction in the Wild", In Proceedings of the European Conference on Computer Vision (ECCV), 2024. (oral presentation) [paper] [github]

[6] Donggyun Kim, Jinwoo Kim, Seongwoong Cho, Chong Luo, Seunghoon Hong, "Universal Few-shot Learning of Dense Prediction Tasks with Visual Token Matching", In Proceedings of the International Conference on Learning Representations (ICLR), 2023. (notable-top-5%, oral presentation) [paper] [github]

[7] Donggyun Kim, Seongwoong Cho, Wonkwang Lee, Seunghoon Hong, "Multi-Task Neural Processes", In Proceedings of the International Conference on Learning Representations (ICLR), 2022. [paper] [github]

[8] Wonkwang Lee, Donggyun Kim, Seunghoon Hong, Honglak Lee, "High-Fidelity Synthesis with Disentangled Representation", In Proceedings of the European Conference on Computer Vision (ECCV), 2020. [paper] [github]

Visit Our Lab: https://vllab.kaist.ac.kr