CV

Download CV as PDF

Education

• Ph.D. in Electronic Information (Artificial Intelligence), School of Computer Science and Engineering, Shanghai Jiao Tong University, 2025 – 2030 (expected)
  • Advisor: Prof. Pengfei Liu
• B.S. in E-Commerce and Law, International School, Beijing University of Posts and Telecommunications, 2021 – 2025
  • National Scholarship (2021); University Second-Class Scholarship and Merit Student (2022)

Work Experience

• Oct 2025 – Sep 2026: Large Language Model Algorithm Intern, Alibaba Group, Taobao & Tmall Group, Business Technology Team
  • Conducted Agent-RL research on fine-grained reward modeling and credit assignment for complex instruction following, leading to the first-author COLM 2026 paper Rubrics to Tokens.
  • Built an end-to-end distributed training pipeline on ROLL (data construction, token-level discriminator training/inference, advantage estimation, policy updates, automated evaluation) supporting multiple Qwen and Llama configurations.
  • Drove RTT’s rollout into the e-commerce data-analysis Agent loop with a hybrid reward (rule signals + expert rubrics + LLM-as-Judge) localized onto key content tokens, reasoning steps, and tool behaviors.
• May – Sep 2023: Algorithm Intern, Lingxin Intelligence
  • Contributed to CharacterGLM, an empathetic persona-consistent model released jointly with Tsinghua’s CoAI Lab; built a large-scale distillation annotation pipeline over frontier LLMs to synthesize, clean, and label personalized dialogue data.

Awards

• First Prize, Beijing Division, 13th and 14th Lanqiao Cup (Software, Group A), 2022 / 2023
• First Prize, Beijing Division, China International College Students’ Innovation Competition (“Painting Music with Colors”), 2024

Publications

MOPD-Router: Rethinking Teacher Routing in Multi-Teacher On-Policy Distillation

Xu, T., Zheng, Y., Zhang, Z., Yu, Y., Ma, C., Zhu, J., Wu, L., Ye, L., Liu, P., Dong, B., Zhu, H., Huang, R., & Yu, G. (2026). "MOPD-Router: Rethinking Teacher Routing in Multi-Teacher On-Policy Distillation." arXiv:2609.30837.

PAC: Progress-Augmented Advantage Curriculum for Multi-Task Reinforcement Learning of LLMs

Yu, Y., Zheng, Y., Zhang, Z., Xu, T., Ma, C., Zhu, J., Liu, J., Deng, X., Dong, B., Zhu, H., & Huang, R. (2026). "PAC: Progress-Augmented Advantage Curriculum for Multi-Task Reinforcement Learning of LLMs." Accepted at EMNLP 2026. arXiv:2608.30528.

ResearcherBench: Evaluating Deep AI Research Systems on the Frontiers of Scientific Inquiry

Xu, T., Lu, P., Ye, L., Hu, X., & Liu, P. (2025). "ResearcherBench: Evaluating Deep AI Research Systems on the Frontiers of Scientific Inquiry." Accepted at COLM 2026. arXiv:2507.16280.

Rubrics to Tokens: Bridging Response-level Rubrics and Token-level Rewards in Instruction Following Tasks

Xu, T., Zheng, Y., Lu, P., Ye, L., Wu, Y., Zhang, Z., Yu, Y., Ma, C., Zhu, J., Liu, P., Dong, B., Zhu, H., Huang, R., & Yu, G. (2026). "Rubrics to Tokens: Bridging Response-level Rubrics and Token-level Rewards in Instruction Following Tasks." Accepted at COLM 2026. arXiv:2604.02795.

ContextBudget: Budget-Aware Context Management for Long-Horizon Search Agents

Wu, Y., Zheng, Y., Xu, T., Zhang, Z., Yu, Y., Zhu, J., Ma, C., Lin, B., Dong, B., Zhu, H., Huang, R., & Yu, G. (2026). "ContextBudget: Budget-Aware Context Management for Long-Horizon Search Agents." Accepted at COLM 2026. arXiv:2604.01664.

Interaction as Intelligence: A Transparent, Fine-Grained Multi-Agent Deep Research System for Human-Agent Collaboration

Ye, L., Cai, X., Wang, X., Wang, J., Zheng, Y., Hu, X., Xu, T., Fu, D., Lu, P., Nan, Y., Su, J., Wu, Y., & Liu, P. (2026). "Interaction as Intelligence: A Transparent, Fine-Grained Multi-Agent Deep Research System for Human-Agent Collaboration." In CHI Conference on Human Factors in Computing Systems (Extended Abstracts).

SkillRouter: Skill Routing for LLM Agents at Scale

Zheng, Y., Zhang, Z., Ma, C., Yu, Y., Zhu, J., Wu, Y., Xu, T., Dong, B., Zhu, H., Huang, R., & Yu, G. (2026). "SkillRouter: Skill Routing for LLM Agents at Scale." Accepted at COLM 2026. arXiv:2603.22455.

InnovatorBench: Evaluating Agents’ Ability to Conduct Innovative LLM Research

Wu, Y., Fu, D., Si, W., Huang, Z., Jiang, M., Li, K., Xia, S., Sun, J., Xu, T., Hu, X., Lu, P., Cai, X., Ye, L., Zhu, W., Xiao, Y., & Liu, P. (2025). "InnovatorBench: Evaluating Agents' Ability to Conduct Innovative LLM Research." Accepted at ICLR 2026. arXiv:2510.27598.

AgencyBench: Benchmarking the Frontiers of Autonomous Agents in 1M-Token Real-World Contexts

Li, K., Shi, J., Xiao, Y., Jiang, M., Sun, J., Wu, Y., Fu, D., Xia, S., Cai, X., Xu, T., Si, W., Li, W., Wang, D., & Liu, P. (2026). "AgencyBench: Benchmarking the Frontiers of Autonomous Agents in 1M-Token Real-World Contexts." In ACL 2026. arXiv:2601.11044.

KG-Adapter: Enabling Knowledge Graph Integration in Large Language Models through Parameter-Efficient Fine-Tuning

Tian, S., Luo, Y., Xu, T., Yuan, C., Jiang, H., Wei, C., & Wang, X. (2024). "KG-Adapter: Enabling Knowledge Graph Integration in Large Language Models through Parameter-Efficient Fine-Tuning." In Findings of the Association for Computational Linguistics: ACL 2024, pages 3813–3828.

Mozart’s Touch: A Lightweight Multi-modal Music Generation Framework Based on Pre-Trained Large Models

Li, J., Xu, T., Chen, X., Yao, X., & Liu, S. (2024). "Mozart's Touch: A Lightweight Multi-modal Music Generation Framework Based on Pre-Trained Large Models." In International Conference on AI-generated Content (AIGC). arXiv:2405.02801.