I am a tenure-track assistant professor in the Department of Computer Science at Illinois Institute of Technology (Illinois Tech). I earned my Ph.D. in Computer Science from the University of Pittsburgh, advised by Dr. Youtao Zhang and Dr. Xulong Tang. I received my M.S. in Telecommunications from the University of Maryland, College Park, and my B.E. in Electrical Engineering and Automation from Beihang University, China.

My research develops full-stack techniques across algorithms, runtimes, and hardware to make AI systems more efficient and reliable. I pursue two connected goals: improving the systems that execute graph-based AI (Systems for AI) and using graph learning to guide efficient and reliable system design (AI for AI Systems). My current work spans three directions:

• Efficient and reliable deep graph learning. I develop algorithms and systems that reduce the training and inference costs of graph learning, from graph matching to temporal graph neural networks. Alongside these optimizations, I study how adversarial inputs affect graph models and their evolving states, with the goal of improving both efficiency and robustness.
• Efficient graph-infused LLMs and agentic systems. I study how graphs can organize external knowledge and experience without introducing excessive construction, retrieval, and execution costs. In GraphRAG, I investigate offline graph compression and adaptive online expansion. In agent memory, I study how to refine and filter past experience so agents can reuse efficient workflows.
• Graph learning for efficient and reliable system design. I investigate how learning can guide decisions about which computation to perform, reuse, or revisit. Building on my earlier work in learning-guided training control and graph-based reinforcement learning for quantum computing, I am extending this approach to selective error recovery in multi-agent systems and efficient image and video generation workflows.

I focus primarily on the first two directions, while developing the third as a complementary extension of my work on learning-guided system optimization.

CV · Email · Google Scholar · LinkedIn

News

• [10/2026] Our recent preprints study input filtering, parallel state reconstruction, and multi-GPU training for temporal graph neural networks. See RAFT, AnchorRM, and our dependency-aware multi-GPU training framework below.
• [08/2025] I joined Illinois Tech as a tenure-track assistant professor of Computer Science.
• [05/2025] Our paper on MemFreezing was accepted to ICML 2025.
• [01/2025] Our papers were accepted to ISCA 2025, ASPLOS 2025, and ICLR 2025.
• [08/2023] Our paper on FlexGM was accepted to ICCD 2023.
• [01/2023] Our paper on SmartFRZ was accepted to ICLR 2023 as a Spotlight.
• [10/2022] Our paper on CEGMA was accepted to HPCA 2023.
• [08/2022] Our paper on ImageArg was accepted to ArgMining 2022.
• [06/2022] Our paper on segmented quantization for graph neural networks was accepted to CCF Transactions on High Performance Computing.

Publications

* Equal contribution.

• Yuting Wu, Mingxing Wang, Xiu Jin, Yue Dai, Liang Liu, Wenjie Yuan, Yong Ye, Zhengkang Zhu, Lichuan Gu, Youtao Zhang. 2026. AgroPatch AutoStrike for Adversarial Patch Attacks on Plant Disease Detection Models. iScience, 29(8).

• Yue Dai*, Liang Liu*, Xulong Tang, Youtao Zhang, Jun Yang. 2025. MemFreezing: A Novel Adversarial Attack on Temporal Graph Neural Networks under Limited Future Knowledge. ICML 2025.

• Yue Dai, Xulong Tang, Youtao Zhang. 2025. Cascade: A Dependency-Aware Efficient Training Framework for Temporal Graph Neural Network. ASPLOS 2025.

• Yingheng Li, Yue Dai, Aditya Pawar, Rongchao Dong, Jun Yang, Youtao Zhang, Xulong Tang. 2025. Reinforcement Learning-Guided Graph State Generation in Photonic Quantum Computers. ISCA 2025.

• Sheng Li, Qitao Tan, Yue Dai, Zhenglun Kong, Tianyu Wang, Jun Liu, Ao Li, Ninghao Liu, Yufei Ding, Xulong Tang, Geng Yuan. 2025. Mutual Effort for Efficiency: A Similarity-based Token Pruning for Vision Transformers in Self-Supervised Learning. ICLR 2025.

• Yue Dai, Xulong Tang, Youtao Zhang. 2023. FlexGM: An Adaptive Runtime System to Accelerate Graph Matching Networks on GPUs. ICCD 2023.

• Sheng Li*, Geng Yuan*, Yue Dai*, Youtao Zhang, Yanzhi Wang, Xulong Tang. 2023. SmartFRZ: An Efficient Training Framework Using Attention-Based Layer Freezing. ICLR 2023, Spotlight.

• Yue Dai, Youtao Zhang, Xulong Tang. 2023. CEGMA: Coordinated Elastic Graph Matching Acceleration for Graph Matching Networks. HPCA 2023.

• Yue Dai, Xulong Tang, Youtao Zhang. 2022. An Efficient Segmented Quantization for Graph Neural Networks. CCF Transactions on High Performance Computing, 4(4), 461–473.

• Zhexiong Liu*, Meiqi Guo*, Yue Dai*, Diane Litman. 2022. ImageArg: A Multi-Modal Tweet Dataset for Image Persuasiveness Mining. ArgMining 2022 at COLING, Best Paper Award.

• Justin Brody, Samuel Barham, Yue Dai, Christopher Maxey, Donald Perlis, David Sekora, Jared Shamwell. 2016. Reasoning with Grounded Self-Symbols for Human-Robot Interaction. AAAI Fall Symposium Series.

Preprints

• Haroon Gharwi, Kai Shu, Yue Dai. 2026. AnchorRM: Parallel Intra-Batch State Reconstruction with Anchor-Based Residual Memory for Temporal Graph Neural Networks. arXiv:2610.27558.

• Yue Dai, Sarah Zhou, Youtao Zhang. 2026. RAFT: Redundancy-Aware Input Filtering for Temporal Graph Neural Network Training. arXiv:2610.24875.

• Yang Zhao*, Yue Dai*, Liang Qiao, Buxin Tu, Bo Liu, Sen Zhang, Yi Zhang, Xiaoyu Hao, Jun Shi, Junshi Chen, Hong An. 2026. A Dependency-Aware Multi-GPU Training Framework for Temporal Graph Neural Networks. arXiv:2610.15730.

• Haroon Gharwi, Yue Dai, Kai Shu. 2026. StateFlow: Dual-State Recurrent Modeling for Long-Horizon Time Series Forecasting. arXiv:2607.00197.

• Sheng Li, Yang Sui, Junhao Ran, Bo Yuan, Yue Dai, Xulong Tang. 2026. Temporal Aware Pruning for Efficient Diffusion-based Video Generation. arXiv:2605.17837.

• Sheng Li, Yang Sui, Yue Wu, Zhuoran Song, Bo Yuan, Xulong Tang, Yue Dai. 2026. Accelerating 3D Gaussian Splatting using Tensor Cores. arXiv:2605.17855.

• Tianyu Wang, Sheng Li, Bingyao Li, Yue Dai, Ao Li, Geng Yuan, Yufei Ding, Youtao Zhang, Xulong Tang. 2024. Improving GPU Multi-Tenancy through Dynamic Multi-Instance GPU Reconfiguration. arXiv:2407.13126.

• Sheng Li, Geng Yuan, Yawen Wu, Yue Dai, Chao Wu, Alex K. Jones, Jingtong Hu, Yanzhi Wang, Xulong Tang. 2024. EdgeOL: Efficient In-situ Online Learning on Edge Devices. arXiv:2401.16694.