Yikun Ban 班义琨
I am a tenure-track associate professor in the School of Computer Science and Engineering at Beihang University and a member of the State Key Laboratory of Software Development Environment. Previously, I was a postdoc and obtained my Ph.D. degree at Computer Science, University of Illinois at Urbana-Champaign. Prior to this, I obtained my Master's degree from EECS, Peking University and bachelor's degree from Wuhan University.
I am interested in principled algorithms in the space of reinforcement learning and deep learning, to solve real-world sequential decision-making problems. Current research topics:
- Reinforcement Learning with Human Feedback
- Multi-Agent Reinforcement Learning
- Ensemble Learning of LLMs
News
- [2025.5] Welcome to check our survey! A Survey on LLM Ensemble.
- [2025.5] NeurIPS 2025 Spotlight! Transformer Copilot: Introduces the new concept of a "Mistake Log" and a novel paradigm for LLM fine-tuning.
- [2025.5] NeurIPS 2025! SamS: Proposes the new problem of dynamic sample scheduling in preference optimization for LLMs, together with an RL-based solution.
- [2025.5] "LLM-Forest" is accepted by ACL 2025 Findings, a new prompt-based LLM Ensemble Learning approach.
- [2024.12] "PageRank Bandit" is accepted by NeurIPS 2024, in which we first use bandit perspective to solve link prediction.
- [2024.12] "Robust Neural Contextual Bandit" is accepted by NeurIPS 2024, in which we remove the Positive Definite assumption for NTK Matrix.
Preprint (* Equal Contribution, # Corresponding)
Selected Publications (* Equal Contribution, # Corresponding)
Transformer Copilot: Learning from The Mistake Log in LLM Fine-tuning
Thirty-ninth Conference on Neural Information Processing Systems (NeurIPS'25, Spotlight)
Adaptive Sample Scheduling for Direct Preference Optimization
Thirty-ninth Conference on Neural Information Processing Systems (NeurIPS'25)
LLM-Forest: Ensemble Learning of LLMs with Graph-Augmented Prompts for Data Imputation
The 63rd Annual Meeting of the Association for Computational Linguistics, Findings (ACL'25)
Adaptive Sampling-based Dynamic Graph Learning for Information Diffusion Prediction
ACM Transactions on Information Systems (TOIS, 2025)
GCL-OT: Graph Contrastive Learning with Optimal Transport for Heterophilic Text-Attributed Graphs
AAAI Conference on Artificial Intelligence (AAAI'26)
[To Appear]
Can Graph Neural Networks Learn Language with Extremely Weak Text Supervision?
The 63rd Annual Meeting of the Association for Computational Linguistics, Main (ACL'25)
Robust Neural Contextual Bandit against Adversarial Corruptions
Thirty-eighth Conference on Neural Information Processing Systems (NeurIPS'24)
PageRank Bandits for Link Prediction
Thirty-eighth Conference on Neural Information Processing Systems (NeurIPS'24)
Meta Clustering of Neural Bandits
ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD'24)
Neural Exploitation and Exploration of Contextual Bandits
Journal of Machine Learning Research (JMLR, To Appear)
Neural Contextual Bandits for Personalized Recommendation
The Web Conference, Tutorial (WWW'24)
Neural Active Learning Beyond Bandits
International Conference on Learning Representations (ICLR'24)
Contextual Bandits with Online Neural Regression
International Conference on Learning Representations (ICLR'24)
Meta-Learning with Neural Bandit Scheduler
Thirty-seventh Conference on Neural Information Processing Systems (NeurIPS'23)
Graph Neural Bandits
ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD'23)
Improved Algorithms for Neural Active Learning
Thirty-sixth Conference on Neural Information Processing Systems (NeurIPS'22)
DISCO: Comprehensive and Explainable Disinformation Detection
ACM International Conference on Information and Knowledge Management (CIKM'22, Demo Track)
Neural Bandit with Arm Group Graph
ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD'22)
EE-Net: Exploitation-Exploration Neural Networks in Contextual Bandits
International Conference on Learning Representations (ICLR'22, Spotlight)
Convolutional Neural Bandit for Visual-aware Recommendation
Preprint: ArXiv:2107.07438
Multi-Facet Contextual Bandits: A Neural Network Perspective
ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD'21)
Local Clustering in Contextual Multi-Armed Bandits
The Web Conference (WWW'21)
Generic Outlier Detection in Multi-Armed Bandit
ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD'20)
No Place to Hide: Catching Fraudulent Entities in Tensors
The Web Conference (WWW'19)
Education
Aug 2019 – Jun 2024
Ph.D., Computer Science, Advised by Jingrui He and Hanghang Tong
University of Illinois at Urbana-Champaign, Illinois, US
Aug 2016 – Jul 2019
M.S., Computer Science
Peking University, Beijing, China
Aug 2012 – Jul 2016
B.S., School of Software Engineering
Wuhan University, Wuhan, China