CUHK AI4Science Team Recruitment: Scientific AGI, Molecular Design, and Scientific Agents

Introduction

The AI4Science team at The Chinese University of Hong Kong is recruiting students and researchers interested in Scientific AGI, AI4Science, generative AI, molecular design, scientific agents, embodied intelligence, and automated laboratories.

This article reorganizes the original recruitment information into a clearer English version for applicants who want to understand the team’s research direction, available positions, application expectations, and research environment. The original structure is preserved, but the wording has been made smoother and easier to read.

Source note: The original article was published on BAAI Hub and states that the content was sourced from WeChat. The body includes one relevant campus/team-related image. Banner ads, site decoration images, and back-to-top UI icons were excluded.

Team Overview

Prof. Pheng Ann Heng’s group in the Department of Computer Science and Engineering at The Chinese University of Hong Kong, together with OTeam, focuses on Scientific Artificial General Intelligence, or Scientific AGI.

The team aims to build a general computational foundation model for AI4Science. Its work explores unified modeling, reasoning, generation, and design across complex scientific systems. Research topics include all-to-all molecular design world models, virtual biological pathway models, Scientific Agents, and self-driving laboratory systems that combine AR/VR with intelligent scientific workflows.

The broader goal is to move AI from being a passive analysis tool toward becoming a core intelligent system that can actively participate in scientific discovery.

The team emphasizes a systematic research path, from foundation models and key algorithms to real scientific problems. Students are expected to develop modeling, reasoning, design, and experimental-loop capabilities in interdisciplinary settings.

The group welcomes students from pharmacy, physics, chemistry, biomedical engineering, computer science, and other related disciplines. It offers an open research atmosphere, strong scientific resources, and substantial computing resources, with nearly one thousand computing cards available. The team also maintains long-term collaborations with leading international AI4Science laboratories, including groups at MIT, Harvard, Stanford, and the Baker Lab.

In addition to basic research, the team has strong industry translation resources and has received nearly one hundred million RMB in industry support. This gives students a platform that combines frontier exploration, interdisciplinary training, and practical scientific application.

Applicants who are interested in Scientific AGI and frontier AI4Science research are encouraged to join and help explore the next generation of AI-driven scientific discovery.

Team Leader

Prof. Pheng Ann Heng is Choh-Ming Li Professor of Computer Science and Engineering at The Chinese University of Hong Kong. He also serves as Director of the Institute of Medical Intelligence and XR at CUHK and Co-Director of the Guangdong-Hong Kong-Macao Joint Laboratory of Human-Machine Intelligence-Synergy Systems.

He received his bachelor’s degree in Computer Science from the National University of Singapore in 1985 and his PhD in Computer Science from Indiana University in
1992. He joined the NUS-JHU Center for Information-enhanced Medicine as a Research Associate in
1992. In 1995, he joined the Department of Computer Science and Engineering at CUHK as an Assistant Professor and was promoted to Professor in 2002.

In 2007, Prof. Heng was awarded the Cheung Kong Scholar Chair Professorship by China’s Ministry of Education. He has also served as department chair and head of the graduate division.

Since 1999, he has directed the Virtual Reality, Visualization and Imaging Research Centre at CUHK. Since 2006, he has also directed the Center for Human-Computer Interaction Technology at the Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences.

His research interests include medical image analysis, artificial intelligence, surgical simulation, visualization, and extended reality. He has published more than 760 papers, with over 80,000 Google Scholar citations and an h-index of
132. He has been named a Highly Cited Researcher by Clarivate and has received recognition from Research.com as a leading figure in computer science in China.

Research Directions

The team supports academic freedom and encourages students to explore research directions that may shape future scientific and AI trends. The current focus includes foundation research in generative models and representation learning, scientific agents, and AI4Science applications.

The following areas are especially relevant, but not exhaustive.

Scientific AI

Deep Generative Foundations and Applications

The team studies fundamental AI methods inspired by physics and related scientific principles. Topics include discrete and continuous diffusion models, other generative modeling methods, and applications such as 3D point clouds, molecule generation, and world models.

This direction is suitable for students who want to work on generative modeling at a foundational level while connecting the methods to scientific problems.

Multimodal Representation Learning

The team studies unified multimodal representation models, including tokenization, modality fusion, and representation-generation integration.

The goal is not only to align different data modalities, but also to make representation learning useful for scientific generation. This includes building more natural and unified ways to generate and reason over scientific structures.

Scientific Agents and Autonomous Discovery

This direction focuses on AI Agents, LLM Agents, and multi-agent systems for scientific tasks. The team explores agents with planning, reasoning, tool use, experimental design, and autonomous optimization capabilities.

Applications include molecular design, protein engineering, automated experiments, literature-based knowledge mining, and scientific hypothesis generation.

The long-term goal is to build a closed loop of:

1. Perception
2. Reasoning
3. Generation
4. Validation
5. Feedback

This loop is intended to accelerate scientific discovery and engineering innovation.

Embodied Scientific Intelligence and Automated Laboratories

This direction studies how virtual intelligence can connect with real experimental systems in AI4Science.

The team focuses on the collaboration between Scientific Agents, robotic experiment platforms, automated laboratories, and intelligent decision-making systems. The intended loop is:

1. Computational design
2. Automated execution
3. Experimental validation
4. Feedback optimization

The goal is to help AI move from virtual reasoning and molecular design into real experimental participation and faster scientific discovery.

AI4Science

Biomolecular World Models

This direction centers on AlloDesign and aims to build a unified foundation model for biological molecular systems.

The research studies how deep generative models and multimodal representation learning can understand molecular structure, function, and interaction. It is not limited to one molecule type. Instead, the same model framework should be able to handle small molecules, nucleic acids, peptides, antibodies, proteins, protein complexes, and other biomolecular systems.

The team aims to build continuous modeling capabilities from molecular representation to functional design. By introducing structure prediction, functional evaluation, and experimental feedback, generated results can be corrected and validated over time.

This may gradually lead to general intelligent design methods for drug discovery, protein engineering, nucleic acid therapeutics, and biomaterial design.

Cell Pathway Models and Aging Research

This direction studies computational modeling of molecular interactions and signaling pathways inside cells.

The key question is how to connect molecular-level representations of proteins, small molecules, and nucleic acids with cellular functional states. The team wants to build predictive models that describe how molecular perturbations lead to cellular responses.

The research pays special attention to aging-related pathways and regulatory mechanisms in disease states. It also explores how AI can understand the effects of molecular interventions on cell fate, signal transduction, and functional decline.

The expected outcome is a new set of computational methods for target discovery, mechanism-of-action analysis, and precise intervention.

AR/VR Virtual Scientists

This direction studies immersive interaction methods for scientific discovery.

The team explores how complex molecular structures, cellular pathways, and model generation processes can be transformed into understandable and interactive 3D research interfaces.

The focus is on collaboration between human scientists and AI Agents. In particular, the team is interested in how interactive visualization and agent systems can support hypothesis generation, experimental design, and result interpretation.

This direction aims to provide new interaction paradigms for future virtual research environments and autonomous scientific discovery systems.

Talent Recruitment

The team is recruiting PhD students for 2027 entry, with around 4 to 6 available places. Postdoctoral positions are open on a rolling basis.

The team also welcomes research assistants, visiting students, and other applicants who want to explore Scientific AI, AI4Science, generative AI, biomolecular design, scientific agents, embodied intelligence, and automated laboratories.

The team believes that genuine research interest and passion are the first drivers of excellent work. Applicants with relevant backgrounds and strong ambition are encouraged to apply. The team will consider each applicant’s research interests, background, and abilities, and then help design a personalized research direction.

If You Are Interested in a PhD Program

The team is looking for potential researchers who are passionate about scientific research, hold themselves to high standards, and have strong academic taste.

The basic expectations for PhD applicants are listed below.

1. Research Experience

For undergraduate students, published papers are not required. However, applicants should have research project experience and should be able to clearly explain their role and contribution in the research process.

For master’s students, it is preferred to have published or submitted papers to top conferences or journals, such as NeurIPS, ICML, CVPR, or ICLR.

Papers available only on arXiv are also welcome. Applicants are encouraged to mark their best one or two works clearly, so the lab can carefully evaluate the quality and research taste behind the work.

1. Bonus Qualifications

Applicants will be especially competitive if they have one or more of the following strengths:

• A strong mathematical or physics background, such as a major in mathematics or physics, or awards in recognized mathematics competitions.
• Strong programming ability, such as a background in computer science or software engineering, participation in programming competitions, or leadership in impactful community projects.
• A high-quality technical blog with long-term updates.
• Active technical contribution records.
• Long-term internship experience in technical roles.

If You Are Interested in Joining as an Intern or Research Assistant

For research assistant roles, the team welcomes candidates with strong interest in research, especially those who:

1. Want to build a solid research foundation before applying for future PhD programs.
2. Have a strong match with the lab’s research directions and are willing to support the team in research, engineering implementation, and data analysis.

Basic Requirements

Applicants should generally have:

• A bachelor’s degree or above.
• A background in mathematics, physics, computer science, or a related field, preferably.
• Good coding ability, such as familiarity with Python, C++, or similar languages.
• Solid engineering implementation skills.
• Strong learning ability and the capacity to quickly master new knowledge and tools.

Bonus Qualifications

The following experience is especially valuable:

• First-author participation in research paper publication.
• Experience developing, optimizing, or maintaining large-scale projects, such as GitHub projects.
• Familiarity with deep learning frameworks such as PyTorch or TensorFlow.
• Experience in research competitions, such as Kaggle or ACM, with awards.

The team also welcomes undergraduate and master’s students as visiting students. Visiting students can participate in the lab’s ongoing frontier research projects.

Basic Requirements for Visiting Students

Visiting student applicants should have:

• Strong motivation to learn and genuine interest in academic research.
• Experience participating in a complete research project, with the ability to explain their contribution and thinking.
• Good English ability, including the ability to read and write academic research materials.

Bonus Qualifications for Visiting Students

Helpful additional strengths include:

• Submitted or published papers in top conferences or journals, not necessarily as first author.
• Strong programming skills.
• Open-source projects or technical blogs.
• Awards in mathematical modeling competitions or other research-related activities.

The team states that it will carefully review all application materials and treat every applicant fairly.

Research Conditions

Computing Resources

The team has strong computing resources through The Chinese University of Hong Kong and the Institute of Medical Intelligence and XR.

In addition, the team works with companies such as Valhalla Technology, Alibaba, Tencent, DP Technology, and BioMap, which also provides significant enterprise-level computing resources.

Exchange and Internship Opportunities

The team maintains long-term collaborations with institutions such as MIT, Harvard, the University of Washington, UCLA, the National University of Singapore, and the University of Cambridge.

PhD students are encouraged to participate in research internships and overseas exchange opportunities, which can support both academic growth and international collaboration.

Academic Exchange

The team encourages students to build collaborations with scholars and top research institutions in the field. It also strongly supports students in attending international academic conferences, including reimbursement support.

For research topics and recent hot areas that students are interested in, the team actively organizes invited talks and promotes relevant online academic exchanges.

Academic Collaboration

The lab supports both online and offline collaboration in AI foundation research and emerging AI research topics. It welcomes collaborators from universities and companies in China and abroad, as long as both sides share compatible research interests.

AI Fundamental Research Collaboration

The lab works closely with research departments in large-model and generative AI companies. Areas of exploration include representation pretraining, diffusion-based generative models, autoregressive and related generative methods, and other frontier AI directions.

AI4Science Project Collaboration

The lab encourages applying AI innovation to life science problems in many forms. It aims to explore the potential of AI in solving complex life science challenges.

The team collaborates with universities and research institutes such as Westlake University, Zhejiang University, Lingang Laboratory, and Shanghai AI Laboratory, as well as leading AI4Science companies such as DP Technology and BioMap. These collaborations cover multiple life science areas, including proteins and small molecules.

How to Apply

Students and researchers interested in the lab can send an email to:

oteam.science@outlook.com

The email subject should follow this format:

School + Name + Application for PhD / Postdoc / Visiting Student / Research Assistant

Applicants should attach:

• A resume or CV.
• A personal statement.
• Other materials that demonstrate their ability, research experience, or technical strengths.

FAQ

What is the CUHK AI4Science team recruiting for?

The team is recruiting 2027-entry PhD students, postdoctoral researchers, research assistants, and visiting students. The research focus includes Scientific AGI, AI4Science, generative AI, molecular design, scientific agents, embodied intelligence, and automated laboratories.

Do undergraduate applicants need published papers?

No. Undergraduate applicants do not need to have published papers, but they should have research project experience. They should also be able to clearly explain what they did, what they contributed, and what they learned from the project.

What backgrounds are suitable for this team?

The team welcomes applicants from computer science, mathematics, physics, chemistry, pharmacy, biomedical engineering, and related fields. Since the research is interdisciplinary, both strong technical ability and genuine interest in scientific problems are important.

What research areas does the team focus on?

Key directions include deep generative models, multimodal representation learning, Scientific Agents, automated laboratories, biomolecular world models, cell pathway modeling, aging research, and AR/VR virtual scientists.

What skills are useful for research assistant applicants?

Strong coding ability is important, especially experience with Python, C++, PyTorch, or TensorFlow. Large-scale engineering experience, research competition experience, GitHub projects, technical blogs, and paper publication experience are also helpful.

Is the team suitable for applicants preparing for future PhD study?

Yes. The research assistant role is especially suitable for candidates who want to build a stronger research foundation before applying for PhD programs. The team also welcomes visiting students who want to participate in frontier research projects.

How should applicants submit materials?

Applicants should email oteam.science@outlook.com with the subject format School + Name + Application for PhD / Postdoc / Visiting Student / Research Assistant. The email should include a CV, a personal statement, and supporting materials that show research and technical ability.

Related Tools

• Google Scholar: Useful for checking publications, citations, and academic research profiles.
• arXiv: A widely used preprint platform for AI, machine learning, and scientific computing research.
• GitHub: Useful for sharing open-source research code, engineering projects, and technical contributions.
• PyTorch: A popular deep learning framework used in AI research and model development.
• TensorFlow: A deep learning platform for building and deploying machine learning models.
• Kaggle: A platform for data science competitions, datasets, notebooks, and applied machine learning practice.

Related Links

• Original BAAI Hub Article: The original Chinese recruitment article.
• The Chinese University of Hong Kong: Official homepage of CUHK.
• CUHK Department of Computer Science and Engineering: Official department website for CUHK CSE.
• CUHK CSE Faculty Page: Official faculty directory for the Department of Computer Science and Engineering.
• Professor Pheng Ann Heng Research Profile: Official CUHK research profile for Prof. Pheng Ann Heng.
• Institute of Medical Intelligence and XR: Official website of the CUHK Institute of Medical Intelligence and XR.
• Prof. Pheng Ann Heng Google Scholar Profile: Publication and citation profile for Prof. Pheng Ann Heng.
• AI for Science in Quantum, Atomistic, and Continuum Systems: A broad AI4Science survey paper for understanding the wider research landscape.

Summary

This article explains the CUHK AI4Science team’s recruitment plan, research directions, application requirements, and research environment. It is most useful for students interested in Scientific AGI, AI4Science, molecular design, scientific agents, and autonomous scientific discovery.

Applicants should pay close attention to the differences between PhD, research assistant, and visiting student expectations. Strong research experience, coding ability, mathematical or scientific background, and clear motivation will all help.

For applicants interested in AI-driven scientific discovery, this team offers a research environment that combines frontier AI methods, interdisciplinary scientific problems, strong computing resources, and international collaboration.