Junyoung Park
Senior AI Researcher
Qualcomm AIjunyoungpark.ml@gmail.com
Latest News
- [8/2025] Our work Routefinder: Towards Foundation Models for Vehicle Routing Problems got accepted to TMLR 2025!
- [7/2025] Our work, RL4CO: an extensive reinforcement learning for combinatorial optimization benchmark got accepted at KDD 2025. Check out our website at https://rl4co.ai4co.org.
- [3/2024] “Recursive Speculative Decoding: Accelerating LLM Inference via Sampling Without Replacement” got accepted to the Workshop on LLM Agents at ICLR 2024.
- [2/2024] Check out the work from our group Direct Alignment of Draft Model for Speculative Decoding with Chat-Fine-Tuned LLMs.
- [1/2024] “Generating Dispatching Rules for the Interrupting Swap-Allowed Blocking Job Shop Problem Using Graph Neural Network and Reinforcement Learning” got accepted to the Journal of Manufacturing Science and Engineering.
- [10/2023] I started as Senior AI Researcher at Qualcomm AI Research to work on efficient LLM!
Education
Ph.D. Korea Advanced Institute of Science & Technology (KAIST)
- Mar 2016 ~ Feb 2023
- Industrial & Systems Engineering
- Advisor: Jinkyoo Park
-
Thesis: Applications of graph neural networks in modeling and decision-making of dynamic networks (Best Dissertation Award)
B.S. Korea Advanced Institute of Science & Technology (KAIST)
- Feb 2011 ~ Feb 2016
- Industrial & Systems Engineering
- Business and Technology Management (Double Major)
- Fully Funded by Korea Scholarship Foundation via National Excellence Scholarship
Selected Papers (*: Equal Contribution)
- Junyoung Park, Dalton Jones, Matt Morse, Raghavv Goel, Mingu Lee, Chris Lott, KeyDiff: Key Similarity-Based KV Cache Eviction for Long-Context LLM Inference in Resource-Constrained Environments, arXiv 2025, [paper]
- Federico Berto, Chuanbo Hua, Nayeli Gast Zepeda, André Hottung, Niels Wouda, Leon Lan, Junyoung Park, Kevin Tierney, Jinkyoo Park, Routefinder: Towards Foundation Models for Vehicle Routing Problems, TMLR 2025. [paper]
- Federico Berto, Chuanbo Hua, Junyoung Park, Laurin Luttmann, et al., RL4CO: an extensive reinforcement learning for combinatorial optimization benchmark, KDD 2025. [Paper][Website]
- Wonsuk Jeon, Manmohan Gagrani, Rishabh Goel, Junyoung Park, Michael Lee, and Christopher Lott, Recursive Speculative Decoding: Accelerating LLM Inference via Sampling Without Replacement, arXiv 2024. [paper]
- Mukul Gagrani, Raghavv Goel, Wonseok Jeon, Junyoung Park, Mingu Lee, Christopher Lott, On Speculative Decoding for Multimodal Large Language Models, arXiv 2024. [paper]
- Raghavv Goel, Mukul Gagrani, Wonseok Jeon, Junyoung Park, Mingu Lee, Christopher Lott, Direct Alignment of Draft Model for Speculative Decoding with Chat-Fine-Tuned LLMs, arXiv 2024. [paper]
- Vivian Wen Hui Wong, Sang Hun Kim, Junyoung Park, Jinkyoo Park, and Kincho H. Law, Generating Dispatching Rules for the Interrupting Swap-Allowed Blocking Job Shop Scheduling Problem Using Graph Neural Network and Reinforcement Learning. MSEC 2023.
- Junyoung Park*, Minjun Kim*, and Jinkyoo Park. Neuro CROSS Exchange: Learning to CROSS Exchange to Solve Realistic Vehicle Routing Problems. ICLR 2023. [paper]
- Junyoung Park, Changhyun Kwon, Jinkyoo Park, Learn to solve the min-max multiple traveling salesmen problem with reinforcement learning. AAMAS 2023.
- Junyoung Park, Federico Berto, Arec Jamgochian, Mykel J. Kochenderfer, and Jinkyoo Park, FOCA: First-order Context-based Adaptation for Generalizing to New Dynamical Systems, arXiv 2023.
- Haewon Jung, Junyoung Park, and Jinkyoo Park, Learning context-aware adaptive solvers to accelerate convex quadratic programming, arXiv 2023. [paper]
- Minsu Kim, Junyoung Park, and Jinkyoo Park, Sym-NCO: Leveraging Symmetricity for Neural Combinatorial Optimization, NeurIPS 2022. [paper] [code]
- Junyoung Park, Jinhyun Choo, and Jinkyoo Park, Convergent Graph Solvers, ICLR 2022. [paper] [code]
- Junyoung Park, Jaehyeong Chun, Sang Hun Kim, Youngkook Kim, and Jinkyoo Park, Learning to schedule job-shop problems: representation and policy learning using graph neural network and reinforcement learning, International Journal of Production Research (IJPR) 2021 (IF = 8.568) Top-cited article published in 2021/22 (certificate) [paper]
- Michael Poli, Stefano Massaroli, Junyoung Park, Atsushi Yamashita, Hajime Asama, and Jinkyoo Park, Graph neural ordinary differential equations, arXiv 2019. [paper] [code]
- Seongcheol Woo, Junyoung Park, Jinkyoo Park, and Lance Manuel, Wind field-based short-term turbine response forecasting by stacked dilated convolutional LSTMs, IEEE Transactions on Sustainable Energy 2019 (IF = 9).
- Junyoung Park, Jinkyoo Park, Physics-induced graph neural network: An application to wind-farm power estimation, Energy 2019 (IF = 8.857) [paper] [code] [slides]