About Me

Hi, I am Yikuan Hu, a researcher at Lamda Group, Nanjing University. My research centers on the reasoning of large language models. I focus on achieving low-cost, high-efficiency optimization without sacrificing performance, with the goal of improving model behavior in complex decision-making and high-precision reasoning scenarios.

Selected Publications

ReMindRAG: Low-Cost LLM-Guided Knowledge Graph Traversal for Efficient RAG

Yikuan Hu*, Jifeng Zhu*, Lanrui Tang, Chen Huang

Advances in Neural Information Processing Systems (NeurIPS), 2025

Paper Code

ASTRO: Automatic Strategy Optimization For Non-Cooperative Dialogues

Yikuan Hu, Chen Huang, Wenqiang Lei

Findings of the Association for Computational Linguistics (ACL), 2025

Paper Code

→ View all publications

News

  • [2026.03] 🎉 Paper accepted at MAR Workshop at CVPR 2026: OrigamiBench: An Interactive Environment to Synthesize Flat-Foldable Origamis.
  • [2026.03] 🎉 Paper accepted at LMReasoning Workshop at AAAI 2026: Abductive State Grounding For Neuro-Symbolic Reinforcement Learning.
  • [2025.12] 🎉 Paper accepted at NeurIPS 2025: ReMindRAG: Low-Cost LLM-Guided Knowledge Graph Traversal for Efficient RAG.
  • [2025.07] 🎉 Paper accepted at Findings of ACL 2025: ASTRO: Automatic Strategy Optimization For Non-Cooperative Dialogues.
  • [2024.10] 🎉 Paper accepted at TRB 2025: Temporal-IRL: Modeling Port Congestion and Berth Scheduling with Inverse Reinforcement Learning.
  • [2024.07] 🌍 Visiting scholar at MIT Center for Transportation & Logistics (MIT CTL), Cambridge, MA.
  • [2024.05] 🏆 Gold Award at the ICPC Kunming Invitational Programming Contest.