About me

Hi! I’m Xueqi Cheng, a Ph.D. student in Computer Science at Florida State University, advised by Dr. Yushun Dong in the Responsible AI (RAI) Lab. I have published first-author papers at top venues including NeurIPS, KDD (Oral), and WSDM, and co-authored papers at ICLR and in journals including ACM Computing Surveys, IEEE TKDE, and ACM TIST. I have conducted industrial research as a research intern at Nokia Applied Research and AT&T Labs. I have received awards including the Naaman Franklin Faile Jr. Graduate Fellowship, the Osher Lifelong Learning Institute Scholarship, the IBM PhD Fellowship, and the WSDM NSF Travel Award, and I serve as a reviewer for venues such as NeurIPS, ICML, KDD, and WWW.

Feel free to drop me an Email if you are interested in collaboration!

Research Interests

• Agentic AI: agent distillation, which transfers the capabilities of large LLM-based agents into smaller and more efficient models, and model routing, which dispatches each query to the most suitable model.
• LLM Compression: making LLMs cheaper to deploy and serve through knowledge distillation, pruning, quantization, and related techniques.
• Multimodal Models: improving the capability, efficiency, and reliability of multimodal large language models (MLLMs).
• AI for Social Good and Real-World Applications: applying AI to societally important problems, including social network analysis and civil and infrastructure engineering.

Selected Publications

^* Equal contribution. See the full publication list.

News

• [09/2026] 🎉 Our paper LatentRouter: Can We Choose the Right Multimodal Large Language Model Before Seeing Its Answer? has been accepted at NeurIPS’26!
• [09/2026] 🎉 Our survey Towards Trustworthy Retrieval Augmented Generation for Large Language Models: A Survey has been published in ACM Computing Surveys!
• [06/2026] 📄 Our preprint Adverse Online Social Interactions: A Multi-Level Evolutionary Analysis of Local Patterns, Diffusion, and Community Disruption is now available online!
• [06/2026] 📄 Our preprint A Nationwide Benchmark for Wildfire Initial Attack Failure Prediction with Public Environmental Data is now available online!
• [05/2026] 🎯 Excited to join Nokia Applied Research as a research intern working on agent distillation!
• [05/2026] 📄 Our preprint ReAD: Reinforcement-Guided Capability Distillation for Large Language Models is now available online!
• [05/2026] 📄 Our preprint LatentRouter: Can We Choose the Right Multimodal Large Language Model Before Seeing Its Answer? is now available online!
• [05/2026] 📄 Our preprint SOMA: Efficient Multi-turn LLM Serving via Small Language Model is now available online!