Publications

Research Interests

• General AI Co-Scientist (SAGA, ARIEL, DrugSAGE, etc.)
• Benchmarking and Evaluations (OSWorld-Science, SciAgentArena, LogicVista, etc.)
• Foundation Models for Biomedical Research (BAITSAO, UNICORN, spEMO, etc.)
• Large Multi-Modal and Reasoning Models (Occamy, TeamPath, RareDx, etc.)

Selected Publications

Underline means these authors contribute equally to this work.

^*: To whom the correspondence should be addressed.

• Accio Lab at Alibaba. Occamy-1.0: Open Pareto-frontier 35B Intelligence for Co-work. Tech Report, 2026 (HuggingFace Daily Papers).
• Du Yuanqi, Yu Botao, Liu Tianyu, et al. Accelerating scientific discovery with autonomous goal-evolving agents. Nature Machine Intelligence, 2027 (in press).
• Liu Tianyu, Huang Tinglin, Ding Tong, Wu Hao, Humphrey Peter, Perincheri Sudhir, Schalper Kurt, Ying Rex, Xu Hua, Zou James, Mahmood Faisal, Zhao Hongyu^*. spEMO: Exploring the Capacity of Foundation Models for Analyzing Spatial Multi-Omic Data. Nature Biomedical Engineering, 2025.
• Liu Tianyu, Chu Tinyi, Luo Xiao, Zhao Hongyu^*. Building A Unified Model for Drug Synergy Analysis Powered by Language Models. Nature Communications, 16.1 (2025): 1-17.
• Liu Tianyu, Huang Tinglin, Lin Yingxin, Ying Rex, Zhao Hongyu^*. UNICORN: Towards universal cellular expression prediction with a multi-task learning framework. Nature Communications 16.1 (2025): 9455.
• Liu, Tianyu and De Brouwer, Edward and Kuo, Tony and Diamant, Nathaniel and Missarova, Alsu and Wang, Hanchen and Hao, Minsheng and Bravo, Hector Corrada and Scalia, Gabriele^* and Regev, Aviv^* and Heimberg Graham^* . Learning multi-cellular representations of single-cell transcriptomics data enables characterization of patient-level disease states. Research in Computational Molecular Biology (RECOMB) and Cell Systems (invited submission), 2026. (oral)
• Liu Tianyu, Li Kexing, Wang Yuge, Li Hongyu, Zhao Hongyu^*. Evaluating the Utilities of Foundation Models in Single-cell Data Analysis. Advanced Science, 2026.
• Liu Tianyu, Zhang Xiangyu, Lin Jiecong, Pinello Luca, Ying Rex, Zhao Hongyu^*. Pre-training Genomic Language Model with Variants for Better Modeling Functional Genomics, NPJ Artificial Intelligence, 2026.
• Liu Tianyu, Chen Tianqi, Zheng Wangjie, Luo Xiao, Chen Yiqun, Zhao Hongyu^*. scELMo: Embeddings from Language Models are Good Learners for Single-cell Data Analysis. Cell Patterns, 2026.
• Liu Tianyu, Xiao Yijia, Luo Xiao, Xu Hua, Zheng Jim Wenjin, Zhao Hongyu^*. Geneverse: A collection of Open-source Multimodal Large Language Models for Genomic and Proteomic Research. The 2024 Conference on Empirical Methods in Natural Language Processing. ACL, 2024. (poster)
• Liu Tianyu, Wang Yuge, Ying Rex, Zhao Hongyu^* . MuSe-GNN: Learning Unified Gene Representation From Multimodal Biological Graph Data. Thirty-seventh Conference on Neural Information Processing Systems. PMLR, 2023. (poster)
• Liu Tianyu^*, Greenberg Grant, Shomorony Ilan. CVQVAE: A Representation Learning based Method for Multi-omics Single Cell Data Integration. Machine Learning in Computational Biology. PMLR, 2022. (Spotlight, top 12%)

Talk

• Tencent Hy, Oct 2026
• BEIHAI Summit, Oct 2026
• THU, Oct 2026
• UMASS, July 2026
• ACL, July 2026
• PSU, Feb 2026
• Rutgers, Feb 2026
• UCSB, Jan 2026
• Duke-NUS, Jan 2026
• Yale BIDS Rising Star Seminar, Dec 2025
• UIUC, Oct 2025
• UW Madison, Oct 2025
• Northwestern University, Oct 2025
• Argonne National Laboratory, Oct 2025
• CGM, Aug 2025
• RECOMB, May 2025
• Chan Zuckerberg Initiative, Mar 2024
• Emory University, Oct 2023
• Novo Nordisk, Oct 2023
• 2nd Open Problems in Single-cell Competition (NeurIPS 2022), Dec 2022
• Second Annual Yale Single Cell Symposium, Oct 2022

Services

As (Associate) Editors or (Area) Chairs:

• NENLP 2025
• ARR
• NeurIPS
• KDD

As reviewers:

• Journals: Nature, Nature Communications, PLOS Comp Bio, BMC Bioinformatics, TNNLS, TALLIP, Communication Biology, Genome Biology, BMC Genomics
• Conferences: NeurIPS, ML4H, MLCB, CVPR, ICLR, ICML, AAAI, AISTATS, ARR
• Competitions: 3rd Open Problems in Single-cell Competition

As organizers:

• NENLP 2025
• AI+Music social activity: NeurIPS 2023.
• LLMs for biomedical research meeting: EMNLP 2024.

As contributors:

• open-source softwares including scGLUE, guidance, Deeprobust, and scVI.