Crab: A Semantics-Aware Checkpoint/Restore Runtime for Agent Sandboxes
Published in arXiv preprint, 2026
Tianyuan Wu*, Chaokun Chang*, Lunxi Cao, Wei Gao, Wei Wang (* Equal contribution).
Publications
You can also find my articles on my Google Scholar profile.
Conference Papers
Weave: Efficient Co-Scheduling for Disaggregated RL Post-Training
Published in USENIX OSDI’26, 2025
Tianyuan Wu, Lunxi Cao, Yining Wei, Wei Gao, Yuheng Zhao, Dakai An Shaopan Xiong, Zhiqiang Lv, Ju Huang, Siran Yang, Yinghao Yu, Jiamang Wang, Lin Qu, and Wei Wang.
RollArt: Scaling Agentic RL Training via Disaggregated Infrastructure
Published in USENIX OSDI’26, 2025
Wei Gao*, Yuheng Zhao*, Tianyuan Wu*, Shaopan Xiong*, Weixun Wang*, Dakai An, Lunxi Cao, Dilxat Muhtar, Zichen Liu, Haizhou Zhao, Ju Huang, Siran Yang, Yongbin Li, Wenbo Su, Jiamang Wang, Lin Qu, Bo Zheng, Wei Wang (* Equal contribution).
AReaL-Hex: Accommodating Asynchronous RL Training over Heterogeneous GPUs
Published in arXiv preprint, 2025
Ran Yan, Youhe Jiang, Tianyuan Wu, Jiaxuan Gao, Zhiyu Mei, Wei Fu, Haohui Mai, Wei Wang, Yi Wu, Binhang Yuan.
RollPacker: Mitigating Long-Tail Rollouts for Fast, Synchronous RL Post-Training
Published in Proceedings of the 23rd USENIX Symposium on Networked Systems Design and Implementation (NSDI’26), 2025
Wei Gao*, Yuheng Zhao*, Dakai An, Tianyuan Wu, Lunxi Cao, Shaopan Xiong, Ju Huang, Weixun Wang, Siran Yang, Wenbo Su, Jiamang Wang, Lin Qu, Bo Zheng, Wei Wang (* Equal contribution).
Virtual Machine NUMA Placement at Scale: Learning the Norm, Shielding the Tail
Published in 9th Annual Conference on Machine Learning and Systems (MLSys’26), 2025
Yibo Zhao*, Tianyuan Wu*, Hui Xue, Qi Chen, Zhenhua Han, Zikai Xu, Yuntai Chang, Rui Gao, Steve Deng, Ray Jui-Hao Chiang, Mingxia Li, Yuqing Yang, Cheng Tan, Fan Yang, Peng Cheng, Yongqiang Xiong, Lili Qiu, Lidong Zhou (* Equal contribution).
Attack of the Bubbles: Straggler-Resilient Pipeline Parallelism for Large Model Training
Published in Proceedings of the 23rd USENIX Symposium on Networked Systems Design and Implementation (NSDI’26), 2025
Tianyuan Wu*, Lunxi Cao*, Hangeng Lu, Xiaoxiao Jiang, Yinghao Yu, Siran Yang, Guodong Yang, Jiamang Wang, Lin Qu, Liping Zhang, Wei Wang. (* Equal contribution) Awarded Outstanding Paper!
Reinforcement Learning Optimization for Large-Scale Learning: An Efficient and User-Friendly Scaling Library
Published in arXiv preprint, 2025
ROLL Team.
Toppings: CPU-Assisted, Rank-Aware Adapter Serving for LLM Inference
Published in 2025 USENIX Annual Technical Conference (USENIX ATC’25), 2025
Suyi Li*, Hanfeng Lu*, Tianyuan Wu, Minchen Yu, Qizhen Weng, Xusheng Chen, Yizhou Shan, Binhang Yuan, Wei Wang (* Equal contribution).
Greyhound: Hunting Fail-Slows in Hybrid-Parallel Training at Scale
Published in 2025 USENIX Annual Technical Conference (USENIX ATC’25), 2025
Tianyuan Wu, Wei Wang, Yinghao Yu, Siran Yang, Wenchao Wu, Qinkai Duan, Guodong Yang, Jiamang Wang, Lin Qu, Liping Zhang.
A Data Optimizer for Region-Aware Self-describing Files in Scientific Computing
Published in 2024 ACM Symposium on Cloud Computing (SoCC’24), 2024
Yanjie Song*, Tianyuan Wu*, Yuanhao Li, Guancheng Li, Yuchen Liu, Shu Yin, Wei Xue, Junchao Wang (* Equal contribution).
Portus: Efficient DNN Checkpointing to Persistent Memory with Zero-Copy
Published in 44th IEEE International Conference on Distributed Computing Systems (ICDCS’24), 2024
Yuanhao Li*, Tianyuan Wu*, Guancheng Li, Yanjie Song, Shu Yin (* Equal contribution).
Journal Articles
Reproducibility: Performance Evaluation of MemXCT on Azure CycleCloud Platform
Published in IEEE Transactions on Parallel and Distributed Systems, 2021
Yuchen Liu*, Yixuan Meng*, Kaiyuan Xu*, Zijun Xu*, Tianyuan Wu*, Yiwei Yang*, Shu Yin* (* All authors contributed equally).