What we believe

"Here's to the crazy ones. The misfits, the rebels, the troublemakers, the round pegs in the square holes, the ones who see things differently. They're not fond of rules, and they have no respect for the status quo. You can quote them, disagree with them, glorify or vilify them. About the only thing you can't do is ignore them. Because they change things - they push the human race forward. And while some may see them as the crazy ones, we see genius. Because the people who are crazy enough to think they can change the world, are the ones who do." --Steve Jobs

"The future depends on some graduate student who is deeply suspicious of everything I have said." -- Geoffrey Hinton

CogAI4Sci team

Our research focuses on developing machine learning methods that enable AI systems to autonomously interact with the world, learn generalizable rules from experience, and drive automatic self-evolution and scientific discovery in efficient and creative ways.

We are recruiting two postdocs in the AI for Science direction and multiple PhD students in either fundamental AI or AI4Sci (see the "JOIN US" tab for details).

Research aims

Research aims

Our team works on three types of automatic AI engines, which we refer to as the three "As".

We use self-evolving AI with improved creativity to help make new discoveries in biomedical sciences

AI for AI that improves AI. We design machine learning models that autonomously plan and conduct interactions with environments and data to learn from experience, aiming for self-evolution. We focus on post-training LLMs and the memories of agentic systems so they can learn from experience instead of relying solely on human data.

We enhance the creativity of generative machine learning models and enable them to propose novel content, focusing on post-training LLMs and the fundamental mechanisms of diffusion.

Team

Team

Abhijeet Sinha (PhD candidate. Previous: IIT, Madras, India)

Arash Lagzian (Research assistant from Sharif University of Technology, Iran.)

Barath Chandran (Intern from IIT Roorkee, India.)

Dianbo Liu (Principal investigator)

Do Huu Dat (Intern from VinUniversity, Vietnam.)

Harshvardhan Saini (Intern. from IIT, India.)

Hengtong Li (PhD candidate co-adv. with Ching-Yu Cheng. Previous: NUS.)

Hongyu He (PhD candidate. Previous: Duke U., USA.)

Hou Hei Lam (Intern from Tsinghua University, China.)

Jing Yu Lim (PhD candidate co-adv. with Tze Yun Leong. Previous: NUS, Singapore)

Jinxiang Xie (Intern. From Nanjing University, China.)

Mike Zhu (PhD candidate at McGill U, Canada, co-adv. with Yue Li.)

Nitin Vetcha (Intern from Indian Institute of Science.)

Peisong Zhang (Master's student at NUS.)

Qiran Zou (PhD candidate co-adv. with Dean Ho. Previous: Tsinghua U, China)

Rushi Shah (PhD candidate. Previous: IIT, Jodhpur, India.)

Samson Yu Bai Jian (PhD candidate co-adv. with Helen Zhou. Previous: NUS)

Shaowen Wang (Intern. From Shanghai Jiao Tong U, China.)

Srinivas Anumasa (Senior postdoctoral researcher. Previous: IIT Hyderabad, India.)

Ting Xu (Postdoc, co-adv. with Ching-Yu Cheng. Previous: USTC, China.)

Tingting Chen (PhD candidate co-adv. with Ching-Yu Cheng. Previous: UNSW, Australia.)

Wenhao Zhao (PhD candidate. Previous: Beihang U, China.)

Xinbao Zou (Intern. from Guangdong University of Foreign Studies, China.)

Yiming Tang (PhD candidate. Previous: Peking U., China.)

Yingtao Zhu (Clinical MBBS student Joint Tsinghua-NUS)

Yudi Wu (Intern. From Zhejiang University, China.)

Yuqian Wang (Intern. From USTC, China.)

Yunyi Lu (Clinical intern. From Jilin University, China.)

Yuxuan Wu (Visiting scholar from Shanghai Maritime University, China.)

Ye Zhang (Visiting faculty member from Tongren Hospital, China. Medicine.)

Yizhen Liao (PhD candidate co-adv. with Roger Foo. Previous: Tsinghua U, China.)

Zarif Ikram (Visiting scholar from Bangladesh University of Southern California.)

Zhaoqian Yao (Intern. From Chinese University of Hong Kong, China.)

Zheng Lin (Intern. From The Hong Kong University of Science and Technology, China.)

Abhishek Das (Intern, Diffusion. From IIT Guwahati, India. Next: back to IIT.)

Ayush Gulati (Intern, Memory. From IIT Jodhpur, India. Next: back to IIT.)

Karaka Prasanth Naidu (Intern. From IIT Dhanbad, India. Next: back to IIT.)

Qifei Wang (Visiting scholar from the Chinese Academy of Sciences. Next: back to CAS.)

Terence Nguang Yi Jie (Intern. From the Singapore Army. Next: undergraduate studies at NUS.)

Xuming Ran (RA. Previous: Chongqing Jiaotong University, China. Next: founded his own startup.)

Jiani Li (Intern from NUS, Singapore. Next: back to NUS.)

Haotian Zheng (Intern from SJTU, China. Next: back to SJTU.)

Ashinee Kesanam (Intern from NIT Karnataka, India. Next: back to India.)

Nirlipta Pande (Intern from BITS Pilani, India. Next: PhD in Europe.)

Anirudh Prabhakaran (Intern from NIT India. Next: Uber.)

Zarif Bin Akhtar (Intern from American International University-Bangladesh. Next: back to Bangladesh.)

Xianrui He (Master's student at NUS. Next: TBD.)

Zexian Wang (Intern from Tsinghua University / University of Michigan. Next: PhD at NUS.)

Zhiwei Xue (Rotation PhD student, NUS. Previous: U. Michigan, USA. Next: PhD at NUS.)

Chengbo Li (Intern from UIUC, USA. Next: Yale University.)

Sankepally Sainath Reddy (Intern from IIIT Raipur, India. Next: back to India.)

Trang Nguyen Ngoc Phuong (Graduate research assistant, previously Mila Canada. Next: PhD at Stanford.)

AmirHossein Alamdar (Intern from Sharif University of Technology, Iran. Next: back to Sharif.)

Aryan Amit Barsainyan (Intern from NITK, India. Next: back to NITK.)

Zhixuan Xiao (Visiting scholar from Tsinghua University, China. Next: back to Tsinghua.)

Jiawei Wu (Intern from Huadong Normal University, China. Next: Rutgers CS PhD.)

Mingyuan Yan (Graduate research assistant from Huadong Normal University, China. Next: NYU ECE PhD.)

Xiaoye Wang (Intern from Harbin Institute of Technology, China. Next: Cambridge University.)

Uma Kadam (Intern from IIIT Guwahati, India. Next: Microsoft.)

Abhinav Sharma (Intern from IIIT Guwahati, India. Next: UMass, USA.)

Manvith Prabhu (Intern from NITK, India. Next: back to India.)

Taoyong Cui (Intern from Tsinghua University, China. Next: back to Tsinghua.)

Anhying Bai (Intern from Tsinghua University, China. Next: back to Tsinghua.)

Zheqi Liu (From Tsinghua University, China. Intern. Next: UCSD.)

Maab Elrashid (From Sudan. Mentee at Mila-Quebec AI Institute with Prof. Yoshua Bengio. Next: Mila.)

Ziqing Mai (From Tsinghua University, China. Intern. Next: back to Tsinghua.)

Zile Yang (Intern from Huazhong University of Science and Technology, China. Next: back to Huazhong.)

Yice Fang (From Tsinghua University, China. Intern. Next: back to Tsinghua.)

Bonaventure F. P. Dossou (Mentee at Mila-Quebec AI Institute. Next: PhD at McGill.)

Rulin Shao (Mentee at MIT/Harvard. Next: graduate student at CMU/Amazon. Now: PhD at UW.)

Loïc Kwate Dassi (Mentee at Mila-Quebec AI Institute. Next: DeepMind London.)

Oussama Boussif (Mentee at Mila. Next: PhD with Yoshua Bengio at Mila.)

Li Huang (Mentee at Harvard. Next: PhD at Tsinghua. Now: faculty at the Chinese Academy of Medical Sciences & Peking Union Medical College.)

James Assiene (Mentee at Mila-Quebec AI Institute. Next: DeepMind London.)

Tianyi Zhang (Mentee at Harvard. Next: PhD at ASU.)

Léna Néhale Ezzine (Mentee at Mila. Next: PhD with Yoshua Bengio at Mila.)

Wisdom d'Almeida (Mentee at Mila-Quebec AI Institute. Next: researcher at Microsoft / PhD at Oxford.)

Jiahe Tian (Mentee at Harvard. Next: graduate student at CMU.)

Ruobin Tao (Mentee in Boston, now at the University of New South Wales, Australia.)

Yuhao Qian (Mentee in Boston. Next: Amazon.)

Pascal Junior Tikeng Notsawo (Mentee at Mila-Quebec AI Institute. Next: PhD at Mila.)

Leyu Dai (Mentee at Harvard. Next: PhD at the University of North Carolina at Chapel Hill.)

Junfeng Zhi (Mentee at Harvard. Next: graduate student at Duke. Now: engineer at Amazon.)

Brice Nanda (Mentee at Mila-Quebec AI Institute. Next: MSc at Mila.)

Yihe Yang (Mentee at Harvard. Next: MSc at CMU.)

Zhuang Ma (Mentee at Harvard. Next: graduate student at CMU.)

Teaching

I am teaching the following courses at National University Of Singapore:

Engineering course EEC4300 Machine Learning: Models and Applications

Engineering course BN5212 Advanced Machine Learning in Biomedical Sciences and Engineering

Medicine course MDG5256 Machine Learning in Medicine

About Dianbo Liu

Before starting CogAI4Sci team, Dianbo Liu was a group leader at the Broad Institute of MIT and Harvard. Prior to the Broad Institute, Dianbo worked as a postdoctoral researcher with Prof. Yoshua Bengio (a Turing Award winner) and led the Humanitarian AI team at the Mila-Quebec AI Institute. This followed his fellowship training and studies in medical informatics at Harvard Medical School. Dianbo earned his PhD from the University of Dundee, Scotland, under the supervision of particle physicist Prof. Timothea Newman. During his doctoral studies,he received the Vest Principal Scholarship from the Massachusetts Institute of Technology (MIT) and was a special graduate student at the MIT Computer Science and Artificial Intelligence Lab. Dianbo also co-founded two start-ups, "GeneTank" and "SecureAILabs," to advance AI applications in sciences during his training.

News

News

  • [Aug 2025] Yizhen Liao and Rushi Shah joined our team as PhD students!
  • [May 2025] Our cell memory work was accepted by Genome Biology.
  • [April 2025] Our distributional shift paper for medical records was accepted by npj Digital Medicine.
  • [March 2025] Our evolutionary GFlowNet paper was accepted by TMLR!
  • [March 2025] Our discrete representation for federated learning paper was accepted by ICML.
  • [Jan 2025] Dr. Ye Zhang joined our team as visiting faculty.
  • [Jan 2025] Samson Yu and Qiran Zou joined our team as PhD students!
  • [June 2024] Our work on self-supervised learning on medical data BarlowTwins-CXR is published in BMC. Congratulations to Haoyue and all co-authors.
  • [Nov 2023] Our exploration of generative models for causal discovery of gene networks Swift-DynGFN was accepted at the NeurIPS Generative Models for Biology workshop. Congratulations to Trang and all co-authors.
  • [Nov 2023] Our large language model physical reasoning task COAT is available.
  • [Jul 2023] We presented our GFlowOut work at ICML 2023 in Hawaii.
  • [Jun 2023] Our two-year effort on Attention Schema will be presented at the NeurIPS InForCog workshop.
  • [Apr 2023] We presented our SAF paper at ICLR 2023 in Kigali, Rwanda.
Join our team

Join our team

Our CogAI4Sci team at NUS is recruiting for the following positions:

  • Two postdoctoral researchers for AI for Science (biomed) research
  • Multiple fully funded PhD students for fundamental machine learning research
  • One fully funded PhD student for AI for Science (biomed) research
  • Research interns and visiting scholars

Our research style: curiosity- and long-term-impact-driven. Team members are self-motivated and like to think big. We are extremely open to different ideas, and team members have a high degree of freedom to explore their interests while we maintain very high research-quality standards.

Who we are looking for: self-motivated candidates with a genuine interest in the scientific questions we are exploring and a passion to change the world.

Our team believes that openness and intellectual freedom are essential to scientific discovery. We warmly welcome candidates who are deeply curious, bold in their thinking, unafraid to challenge convention, and eager to see the world from new perspectives—regardless of their undergraduate institution, gender, religion, race, age, national origin, or disability.

If you are interested, please fill in this form and email me at dianbo at nus dot edu dot sg.

Publications

Publications

For the most up-to-date list of publications, see my Google Scholar profile.

Selected Machine Learning Publications

GUncertainty-Based Extensible Codebook for Discrete Federated Learning in Heterogeneous Data Silos
Zhang, Tianyi, Yu Cao, and Dianbo Liu
ICML 2025
Unsupervised concept discovery mitigates spurious correlations
Arefin, Md Rifat, Yan Zhang, Aristide Baratin, Francesco Locatello, Irina Rish, Dianbo Liu, and Kenji Kawaguchi
ICML 2024
GFlowOut: Dropout with Generative Flow Networks
Dianbo Liu , Moksh Jain, Bonaventure Dossou, Qianli Shen, Salem Lahlou, Anirudh Goyal, Nikolay Malkin, Chris Emezue, Dinghuai Zhang, Nadhir Hassen, Xu Ji, Kenji Kawaguchi, Yoshua Bengio
ICML 2023
Stateful Active Facilitator: Coordination and Environmental Heterogeneity in Cooperative Multi-Agent Reinforcement Learning
Dianbo Liu, Vedant Shah, Oussama Boussif, Anirudh Goyal, Michael Curtis Mozer, Nicolas Heess, Yoshua Bengio.
ICLR 2023
Discrete-Valued Neural Communication
Dianbo Liu, Alex Lamb, Kenji Kawaguchi, Anirudh Goyal, Chen Sun, Michael Curtis Mozer, Yoshua Bengio
NeurIPS 2021

Selected Machine Learning for Biomed Publications

FedWeight: mitigating covariate shift of federated learning on electronic health records data through patient re-weighting
Zhu, He, Jun Bai, Na Li, Xiaoxiao Li, Dianbo Liu, David L. Buckeridge, and Yue Li.
npj Digital Medicine 2025
Machine learning approaches to predicting no-shows in pediatric medical appointments
Dianbo Liu, Won-Yong Shin, Eli Sprecher, Kathleen Conroy, Omar Santiago, Gal Wachtel, Mauricio Santillana
npj Digital Medicine 2022
ENCODE Phase III: Building an Encyclopedia of Candidate cis-Regulatory Elements for Human and Mouse
Jill Moore, Michael J. Purcaro, Bradley E. Bernstein, ... Dianbo Liu, ... Barbara Wold, Ross C. Hardison, et al.
Nature 2020
Integrative construction of regulatory region networks in 127 human reference epigenomes by matrix factorization
Dianbo Liu, Jose Davila-Velderrain, Zhizhuo Zhang, Manolis Kellis, et al.
Nucleic Acids Research 2019
Get in Touch

Contact

Level 13
12 Science Drive 2, Singapore 117549