Yuyang Wu

Ph.D. Student, Carnegie Mellon University

yuyangwu [at] andrew.cmu.edu

Email / Google Scholar / GitHub / X / CV

I am Yuyang (Younger) Wu, 吴雨阳, a Ph.D. student in AI for Science at Carnegie Mellon University, where I am fortunate to be advised by Prof. Olexandr Isayev. I received my M.S. from CMU’s School of Computer Science. My work so far has focused on making LLM reasoning trustworthy and efficient for chemistry and science, from benchmarking chemical errors and cost reasoning to compressing and editing long reasoning traces. Building on this, I am now working toward self-evolving scientific agents: systems that keep improving their own capabilities, strategies, and knowledge from experience to accelerate scientific discovery.

I have also been fortunate to collaborate with Graham Neubig on the AI Science Foundry and cloud lab for autonomous scientific experimentation, Manling Li on efficient LLM reasoning, Yonatan Bisk on trustworthy LLMs for chemistry, and Jian Ma and Xiangliang Zhang on evaluating LLM agents for chemical reasoning. Beyond academia, I am collaborating with NVIDIA on Scientific Agents, and previously worked with OpenAI to improve the chemical capabilities of frontier LLMs, focusing on rigorous chemical reasoning and reliable domain knowledge.

I believe science is the ultimate mission of AI. 科学是AI的终极任务。

我享受充分的睡眠和多交朋友带来的创造力。

news

Sep 25, 2026	Our paper Can Agents Price a Reaction? Evaluating LLMs on Chemical Cost Reasoning (ChemCost) has been accepted to NeurIPS 2026! 🎉
Sep 17, 2026	New preprint on bioRxiv: High-throughput physics-based enzyme engineering!
Aug 01, 2026	Our paper MolecularCanvas: LLM-assisted Small-Molecule Drug Discovery via Structure-Guided Constraints has been accepted to UIST 2026!
Jul 31, 2026	Our paper KGRxn-LLM: Knowledge Graph Enhanced Large Language Models for Molecular Reaction Reasoning has been accepted to BioNLP 2026 (ACL workshop)!
Jul 27, 2026	Our paper VeriLLMed: Interactive Visual Debugging of Medical Large Language Models with Knowledge Graphs has been accepted to VIS 2026!

publications

selected

NeurIPS

Can Agents Price a Reaction? Evaluating LLMs on Chemical Cost Reasoning

Yuyang Wu, Yue Huang, Shuaike Shen, Xujian Wang, Shuhao Zhang, Qiyao Xue, Weichen Liu, Runtian Gao, Jian Ma, Xiangliang Zhang, and Olexandr Isayev

In Advances in Neural Information Processing Systems (NeurIPS), 2026

arXiv Bib Code

@inproceedings{wu2026chemcost,
  title = {Can Agents Price a Reaction? Evaluating LLMs on Chemical Cost Reasoning},
  author = {Wu, Yuyang and Huang, Yue and Shen, Shuaike and Wang, Xujian and Zhang, Shuhao and Xue, Qiyao and Liu, Weichen and Gao, Runtian and Ma, Jian and Zhang, Xiangliang and Isayev, Olexandr},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2026},
}

arXiv

KARA: Efficient Reasoning LLM Serving via Sliding-Window KV Cache Compression

Shen Han^*, Yuyang Wu^*, Junpu Yu, and Olexandr Isayev

arXiv preprint arXiv:2607.01237, 2026

arXiv Bib

@article{han2026kara,
  title = {KARA: Efficient Reasoning LLM Serving via Sliding-Window KV Cache Compression},
  author = {Han, Shen and Wu, Yuyang and Yu, Junpu and Isayev, Olexandr},
  journal = {arXiv preprint arXiv:2607.01237},
  year = {2026},
}

arXiv

CLORE: Content-Level Optimization for Reasoning Efficiency

Yuyang Wu, Qiyao Xue, Guanxing Lu, Weichen Liu, Zihan Wang, Manling Li, and Olexandr Isayev

arXiv preprint arXiv:2605.22211, 2026

arXiv Bib Code

@article{wu2026clore,
  title = {CLORE: Content-Level Optimization for Reasoning Efficiency},
  author = {Wu, Yuyang and Xue, Qiyao and Lu, Guanxing and Liu, Weichen and Wang, Zihan and Li, Manling and Isayev, Olexandr},
  journal = {arXiv preprint arXiv:2605.22211},
  year = {2026},
}

EMNLP

MolErr2Fix: Benchmarking LLM Trustworthiness in Chemistry via Modular Error Detection, Localization, Explanation, and Revision

Yuyang Wu, Jinhui Ye, Shuhao Zhang, Lu Dai, Yonatan Bisk, and Olexandr Isayev

In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP) Oral Presentation , 2025

Bib HTML PDF Code

@inproceedings{wu2025molerr2fix,
  title = {MolErr2Fix: Benchmarking LLM Trustworthiness in Chemistry via Modular Error Detection, Localization, Explanation, and Revision},
  author = {Wu, Yuyang and Ye, Jinhui and Zhang, Shuhao and Dai, Lu and Bisk, Yonatan and Isayev, Olexandr},
  booktitle = {Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP)},
  year = {2025},
  publisher = {Association for Computational Linguistics},
}

others

ECCV

Reasoning Path and Latent State Analysis for Multi-view Visual Spatial Reasoning: A Cognitive Science Perspective

Qiyao Xue, Weichen Liu, Shiqi Wang, Haoming Wang, Yuyang Wu, and Wei Gao

In European Conference on Computer Vision (ECCV), 2026

arXiv Bib Code

@inproceedings{xue2026reasoning,
  title = {Reasoning Path and Latent State Analysis for Multi-view Visual Spatial Reasoning: A Cognitive Science Perspective},
  author = {Xue, Qiyao and Liu, Weichen and Wang, Shiqi and Wang, Haoming and Wu, Yuyang and Gao, Wei},
  booktitle = {European Conference on Computer Vision (ECCV)},
  year = {2026},
}

BioNLP

KGRxn-LLM: Knowledge Graph Enhanced Large Language Models for Molecular Reaction Reasoning

Weichen Liu, Qiyao Xue, Yuyang Wu, Olexandr Isayev, and Natasa Miskov-Zivanov

In Proceedings of the BioNLP Workshop (ACL), 2026

DOI Bib HTML Code

@inproceedings{liu2026kgrxnllm,
  title = {KGRxn-LLM: Knowledge Graph Enhanced Large Language Models for Molecular Reaction Reasoning},
  author = {Liu, Weichen and Xue, Qiyao and Wu, Yuyang and Isayev, Olexandr and Miskov-Zivanov, Natasa},
  booktitle = {Proceedings of the BioNLP Workshop (ACL)},
  pages = {268--281},
  year = {2026},
  publisher = {Association for Computational Linguistics},
  doi = {10.18653/v1/2026.bionlp-1.22},
}

VIS

VeriLLMed: Interactive Visual Debugging of Medical Large Language Models with Knowledge Graphs

Yurui Xiang, Xingyi Mao, Rui Sheng, Zixin Chen, Zelin Zang, Yuyang Wu, Haipeng Zeng, Huamin Qu, Yushi Sun, and Yanna Lin

In IEEE Visualization Conference (VIS), 2026

arXiv Bib

@inproceedings{xiang2026verillmed,
  title = {VeriLLMed: Interactive Visual Debugging of Medical Large Language Models with Knowledge Graphs},
  author = {Xiang, Yurui and Mao, Xingyi and Sheng, Rui and Chen, Zixin and Zang, Zelin and Wu, Yuyang and Zeng, Haipeng and Qu, Huamin and Sun, Yushi and Lin, Yanna},
  booktitle = {IEEE Visualization Conference (VIS)},
  year = {2026},
}

UIST

MolecularCanvas: LLM-assisted Small-Molecule Drug Discovery via Structure-Guided Constraints

Haoyu Dong, Rui Sheng, Shuhao Zhang, Yushi Sun, Dingyang Wu, Hanxiang Chao, Olexandr Isayev, Huamin Qu, Yuyang Wu^†, and Yanna Lin

In ACM Symposium on User Interface Software and Technology (UIST), 2026

arXiv Bib

@inproceedings{dong2026molecularcanvas,
  title = {MolecularCanvas: LLM-assisted Small-Molecule Drug Discovery via Structure-Guided Constraints},
  author = {Dong, Haoyu and Sheng, Rui and Zhang, Shuhao and Sun, Yushi and Wu, Dingyang and Chao, Hanxiang and Isayev, Olexandr and Qu, Huamin and Wu, Yuyang and Lin, Yanna},
  booktitle = {ACM Symposium on User Interface Software and Technology (UIST)},
  year = {2026},
}

bioRxiv

High-throughput physics-based enzyme engineering

Xujian Wang, Xiang Qiu, Yuyang Wu, Taoyu Niu, Shuhao Zhang, Runtian Gao, Ilkwon Cho, Haocheng Tang, Kangdelong Hu, Xiaoguang Lei, Olexandr Isayev, and Junmei Wang

bioRxiv preprint (under review), 2026

DOI Bib

@article{wang2026enzyme,
  title = {High-throughput physics-based enzyme engineering},
  author = {Wang, Xujian and Qiu, Xiang and Wu, Yuyang and Niu, Taoyu and Zhang, Shuhao and Gao, Runtian and Cho, Ilkwon and Tang, Haocheng and Hu, Kangdelong and Lei, Xiaoguang and Isayev, Olexandr and Wang, Junmei},
  journal = {bioRxiv preprint (under review)},
  year = {2026},
  doi = {10.64898/2026.09.15.751901},
}

BIB

Multiview representation learning for identification of novel cancer genes and their causative biological mechanisms

Jianye Yang, Haitao Fu, Feiyang Xue, Menglu Li, Yuyang Wu, Zhanhui Yu, Haohui Luo, Jing Gong, Xiaohui Niu, and Wen Zhang

Briefings in Bioinformatics, 2024

DOI Bib Code

@article{yang2024multiview,
  title = {Multiview representation learning for identification of novel cancer genes and their causative biological mechanisms},
  author = {Yang, Jianye and Fu, Haitao and Xue, Feiyang and Li, Menglu and Wu, Yuyang and Yu, Zhanhui and Luo, Haohui and Gong, Jing and Niu, Xiaohui and Zhang, Wen},
  journal = {Briefings in Bioinformatics},
  volume = {25},
  number = {5},
  pages = {bbae418},
  year = {2024},
  publisher = {Oxford Academic},
  doi = {10.1093/bib/bbae418},
}

BIB

HimGNN: a novel hierarchical molecular graph representation learning framework for property prediction

Shen Han, Haitao Fu, Yuyang Wu, Ganglan Zhao, Zhenyu Song, Feng Huang, Zhongfei Zhang, Shichao Liu, and Wen Zhang

Briefings in Bioinformatics, 2023

DOI Bib Code

@article{han2023himgnn,
  title = {HimGNN: a novel hierarchical molecular graph representation learning framework for property prediction},
  author = {Han, Shen and Fu, Haitao and Wu, Yuyang and Zhao, Ganglan and Song, Zhenyu and Huang, Feng and Zhang, Zhongfei and Liu, Shichao and Zhang, Wen},
  journal = {Briefings in Bioinformatics},
  volume = {24},
  number = {5},
  pages = {bbad305},
  year = {2023},
  publisher = {Oxford Academic},
  doi = {10.1093/bib/bbad305},
}

Experience

Carnegie Mellon University — Ph.D. in AI for Science, advised by Olexandr Isayev (2026 – present)
Carnegie Mellon University, School of Computer Science — M.S. in Automated Science (2024 – 2026)
National University of Singapore, School of Computing — Visiting Student (Summer 2023)
Huazhong Agricultural University — B.S., Zhang Zhidong Advanced Class (2020 – 2024)

Honors

Samuel S. Wang Award
Oral Presentation, EMNLP 2025 — MolErr2Fix

Services

Conference Reviewer: NeurIPS, ICLR, ACL Rolling Review (ARR), EMNLP
Journal Reviewer: American Chemical Society (ACS) journals