Publications

2026

1. Preprint

   DenoiseRL: Bootstrapping Reasoning Models to Recover from Noisy Prefixes

   Caijun Xu, Changyi Xiao, Zhongyuan Peng, and Yixin Cao

   arXiv preprint arXiv:2605.28421, May 2026

   Abs arXiv Bib PDF

   DenoiseRL is a reinforcement learning framework that replaces external supervision with recovery-oriented optimization over failures from weak models. It learns directly from incorrect reasoning traces by conditioning the policy on noisy prefixes and optimizing the on-policy continuation, improving reasoning performance, exploration efficiency, and self-corrective behavior without stronger teacher models or heavily curated difficult datasets.

   @article{xu2026denoiserl,
     title = {{DenoiseRL: Bootstrapping Reasoning Models to Recover from Noisy Prefixes}},
     author = {Xu, Caijun and Xiao, Changyi and Peng, Zhongyuan and Cao, Yixin},
     journal = {arXiv preprint arXiv:2605.28421},
     year = {2026},
     month = may,
     url = {https://arxiv.org/abs/2605.28421},
   }

2. Preprint

   Reinforcement Learning with Conditional Expectation Reward

   Changyi Xiao, Caijun Xu, and Yixin Cao

   arXiv preprint arXiv:2603.10624, Mar 2026

   arXiv Bib PDF Code

   @article{xiao2026conditional,
     title = {{Reinforcement Learning with Conditional Expectation Reward}},
     author = {Xiao, Changyi and Xu, Caijun and Cao, Yixin},
     journal = {arXiv preprint arXiv:2603.10624},
     year = {2026},
     month = mar,
     url = {https://arxiv.org/abs/2603.10624},
   }

3. Preprint

   CoDiQ: Test-Time Scaling for Controllable Difficult Question Generation

   Zhongyuan Peng, Caijun Xu, Changyi Xiao, Shibo Hong, Eli Zhang, Stephen Huang, and Yixin Cao

   arXiv preprint arXiv:2602.01660, Feb 2026

   arXiv Bib PDF

   @article{peng2026codiq,
     title = {{CoDiQ: Test-Time Scaling for Controllable Difficult Question Generation}},
     author = {Peng, Zhongyuan and Xu, Caijun and Xiao, Changyi and Hong, Shibo and Zhang, Eli and Huang, Stephen and Cao, Yixin},
     journal = {arXiv preprint arXiv:2602.01660},
     year = {2026},
     month = feb,
     url = {https://arxiv.org/abs/2602.01660},
   }

4. ACL’26 Findings

   SCALER: Synthetic Scalable Adaptive Learning Environment for Reasoning

   Caijun Xu, Changyi Xiao, Zhongyuan Peng, Xinrun Wang, and Yixin Cao

   In Findings of the Association for Computational Linguistics: ACL 2026, Jan 2026

   arXiv Bib

   @inproceedings{xu2026scaler,
     title = {{SCALER: Synthetic Scalable Adaptive Learning Environment for Reasoning}},
     author = {Xu, Caijun and Xiao, Changyi and Peng, Zhongyuan and Wang, Xinrun and Cao, Yixin},
     booktitle = {Findings of the Association for Computational Linguistics: {ACL} 2026},
     year = {2026},
     month = jan,
     url = {https://arxiv.org/abs/2601.04809},
   }

2024

1. CIKM’24

   Exploring High-Order User Preference with Knowledge Graph for Recommendation

   Caijun Xu, Fuwei Zhang, Zhao Zhang, Fuzhen Zhuang, and Rui Liu

   In Proceedings of the 33rd ACM International Conference on Information and Knowledge Management, Jan 2024

   Bib

   @inproceedings{xu2024exploring,
     title = {{Exploring High-Order User Preference with Knowledge Graph for Recommendation}},
     author = {Xu, Caijun and Zhang, Fuwei and Zhang, Zhao and Zhuang, Fuzhen and Liu, Rui},
     booktitle = {Proceedings of the 33rd {ACM} International Conference on Information and Knowledge Management},
     pages = {4138--4142},
     year = {2024},
     url = {https://doi.org/10.1145/3627673.3679921},
   }