Shouren Wang

About Me

I’m Shouren Wang, a PhD student in Computer Science at Case Western Reserve University, advised by Prof. Xiaotian Han and Prof. Vipin Chaudhary. I am a member of TAPP AI Lab.

Previously, I earned an M.S. in Computer Engineering from New York University, where I was part of the Game Innovation Lab, and a B.E. in Digital Media Technology from Hunan University.

I am actively seeking internship opportunities and would be glad to discuss potential research collaborations.

Research Interests

My current work broadly concerns large language model reasoning, post-training, efficient language models, model architecture, and AI agent systems.

News

Oct 26, 2026	My advisor and co-author, Prof. Vipin Chaudhary, is at EMNLP 2026 in Budapest to present our paper, Demystifying Hybrid Thinking. If you are attending, feel free to connect with him and chat about our work!
Oct 06, 2026	I am at COLM 2026 in San Francisco! I will be presenting our paper Path-Lock Expert. If you are attending, feel free to come say hello and chat! 👋
Oct 04, 2026	Agent Team Work Zone v1.0.0 is officially available on npm! Install it into a project with `npx agent-team-work-zone init`.
Sep 28, 2026	Our preprint How to Loop MoE: Flatten the Experts, Untie the Attention is now available on arXiv! We introduce Foil, which flattens expert layers and unties attention to improve looped MoE models at fixed parameters and compute. 🚀
Aug 24, 2026	My first-author paper Demystifying Hybrid Thinking was accepted to the EMNLP 2026 Main Conference! 🎉
Aug 04, 2026	Our survey Test-Time Scaling in Reasoning LLMs is now available on arXiv! 📄
Jul 24, 2026	I released Agent Team Work Zone, a persistent workspace and management layer for long-lived Claude Code agent teams! 🚀
Jul 08, 2026	My first-author paper Path-Lock Expert was accepted to COLM 2026! 🎉
Jun 30, 2026	My first-author survey A Survey on Efficient Protein Language Models was published in TMLR! 🎉
Jun 15, 2026	Our paper MemTrace is now available on arXiv! 📄
May 21, 2026	Our paper CausalGuard: Conformal Inference under Graph Uncertainty is now available on arXiv! 📄
May 15, 2026	Our survey A Survey on Agent Skills for LLMs is now available on SSRN! 📄
May 13, 2026	Our paper Reliability-Gated Source Anchoring for Continual Test-Time Adaptation is now available on arXiv! 📄

Selected Publications

Path-Lock Expert: Separating Reasoning Mode in Hybrid Thinking via Architecture-Level Separation

Shouren Wang^*, Wang Yang^*, Chuang Ma, Debargha Ganguly, Vikash Singh, Chaoda Song, Xinpeng Li, Xianxuan Long, Vipin Chaudhary, and Xiaotian Han

In Conference on Language Modeling, Oct 2026

arXiv Bib PDF Code

@inproceedings{wang2026pathlock,
  title = {Path-Lock Expert: Separating Reasoning Mode in Hybrid Thinking via Architecture-Level Separation},
  author = {Wang, Shouren and Yang, Wang and Ma, Chuang and Ganguly, Debargha and Singh, Vikash and Song, Chaoda and Li, Xinpeng and Long, Xianxuan and Chaudhary, Vipin and Han, Xiaotian},
  booktitle = {Conference on Language Modeling},
  year = {2026},
  first_public_year = {2026},
  month = oct,
}

Demystifying Hybrid Thinking: Can LLMs Truly Switch Between Think and No-Think?

Shouren Wang^*, Wang Yang^*, Xianxuan Long, Qifan Wang, Vipin Chaudhary, and Xiaotian Han

In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, Oct 2026

Preprint: Oct 2025 · Accepted: EMNLP 2026 Main

arXiv Bib PDF Code

@inproceedings{wang2026demystifying,
  title = {Demystifying Hybrid Thinking: Can LLMs Truly Switch Between Think and No-Think?},
  author = {Wang, Shouren and Yang, Wang and Long, Xianxuan and Wang, Qifan and Chaudhary, Vipin and Han, Xiaotian},
  booktitle = {Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing},
  year = {2026},
  first_public_year = {2025},
  month = oct,
  note = {Preprint: Oct 2025 · Accepted: EMNLP 2026 Main},
}

Mid-Think: Training-Free Intermediate-Budget Reasoning via Token-Level Triggers

Wang Yang^*, Shouren Wang^*, Debargha Ganguly, Xinpeng Li, Chaoda Song, Vikash Singh, Vipin Chaudhary, and Xiaotian Han

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

Bib HTML PDF Code

@inproceedings{yang2026midthink,
  title = {Mid-Think: Training-Free Intermediate-Budget Reasoning via Token-Level Triggers},
  author = {Yang, Wang and Wang, Shouren and Ganguly, Debargha and Li, Xinpeng and Song, Chaoda and Singh, Vikash and Chaudhary, Vipin and Han, Xiaotian},
  booktitle = {Findings of the Association for Computational Linguistics: ACL 2026},
  pages = {6024--6038},
  year = {2026},
  first_public_year = {2026},
  month = jul,
  publisher = {Association for Computational Linguistics},
  doi = {10.18653/v1/2026.findings-acl.299},
}

A Survey on Efficient Protein Language Models

Shouren Wang, Debargha Ganguly, Vinooth Rao Kulkarni, Wang Yang, Zhuoran Qiao, Daniel Blankenberg, Vipin Chaudhary, and Xiaotian Han

Transactions on Machine Learning Research, Jun 2026

Preprint: 2025 · Published: TMLR 2026

Bib HTML PDF Code

@article{wang2026efficientplm,
  title = {A Survey on Efficient Protein Language Models},
  author = {Wang, Shouren and Ganguly, Debargha and Kulkarni, Vinooth Rao and Yang, Wang and Qiao, Zhuoran and Blankenberg, Daniel and Chaudhary, Vipin and Han, Xiaotian},
  journal = {Transactions on Machine Learning Research},
  year = {2026},
  first_public_year = {2025},
  month = jun,
  note = {Preprint: 2025 · Published: TMLR 2026},
}

How to Loop MoE: Flatten the Experts, Untie the Attention

Shouren Wang, Chuang Ma, Mohsen Hariri, Debargha Ganguly, Wang Yang, Xiaoqing Tong, Qianying Liu, Xiaotian Han, and Vipin Chaudhary

arXiv preprint arXiv:2609.35751, Sep 2026

Under Review

arXiv Bib PDF Code

@article{wang2026howtoloopmoe,
  title = {How to Loop {MoE}: Flatten the Experts, Untie the Attention},
  author = {Wang, Shouren and Ma, Chuang and Hariri, Mohsen and Ganguly, Debargha and Yang, Wang and Tong, Xiaoqing and Liu, Qianying and Han, Xiaotian and Chaudhary, Vipin},
  journal = {arXiv preprint arXiv:2609.35751},
  note = {Under Review},
  year = {2026},
  first_public_year = {2026},
  month = sep,
}

Agent Team Work Zone: An Automated, Persistent Workspace for Long-Lived Claude Code Agent Teams

Shouren Wang

arXiv preprint arXiv:2607.22917, Jul 2026

Preprint · Open Source · npm Package

arXiv Bib HTML PDF Code

@article{wang2026atwz,
  title = {Agent Team Work Zone: An Automated, Persistent Workspace for Long-Lived Claude Code Agent Teams},
  author = {Wang, Shouren},
  journal = {arXiv preprint arXiv:2607.22917},
  note = {Preprint · Open Source · npm Package},
  year = {2026},
  first_public_year = {2026},
  month = jul,
}

Service

Reviewer, Conference on Language Modeling (COLM), 2026

Teaching

Teaching Assistant, CSDS 600 - Large Language Models, Prof. Xiaotian Han, Case Western Reserve University (Fall 2024)
Teaching Assistant, CSDS 433 - Database Systems, Prof. Yinghui Wu, Case Western Reserve University (Fall 2026)