Our research aims to advance AI while developing methods that enable new discoveries in biology and medicine. We design machine learning approaches for complex, heterogeneous, and multimodal biomedical data, with an emphasis on models that can generalize beyond the settings observed during training. We also seek to understand when and why AI models fail, using these failure modes to uncover limitations of current methods and guide the development of more robust and generalizable learning principles. Alongside advancing fundamental AI research, we apply these methods to challenging problems in biomedicine, with the goal of revealing novel biological insights.
Advancing machine learning research
Our machine learning research aims to advance the capabilities and our understanding of modern AI systems. We develop methods that enable foundation models to adapt and generalize with limited or no supervision. We study multimodal AI, seeking to understand how representations across different modalities relate and how to effectively align models across different modalities. Across these directions, we investigate the capabilities, limitations, and failure modes of modern AI systems, using these insights to develop more robust and generalizable learning principles.
Example publications:
- Revisiting the Platonic Representation Hypothesis: An Aristotelian View
- PACER: Acyclic Causal Discovery from Large-scale Interventional Data
- Meta-RL Induces Exploration in Language Agents
- When Weak LLMs Speak with Confidence, Preference Alignment Gets Stronger
- HeurekaBench: A Benchmarking Framework for AI Co-scientist
- With Limited Data for Multimodal Alignment, Let the STRUCTURE Guide You
- Weak-to-Strong Generalization under Distribution Shifts
- Large (Vision) Language Models are Unsupervised In-Context Learners
- Let Go of Your Labels with Unsupervised Transfer
- Cross-domain Open-world Discovery
- Fine-grained Classes and How to Find Them
- The Pursuit of Human Labeling: A New Perspective on Unsupervised Learning
- Open-World Semi-Supervised Learning
- Concept Learners for Few-Shot Learning
Advancing biomedical research
We use our methods to unravel new biological insights and work on transformative applications of machine learning in biomedicine. We are especially interested in using our methods for the analysis and understanding of high throughput single-cell data which provides unprecedented resolution to examine cell heterogeneity in tissues and organs. Our methods have been used to annotate global cell atlas consortia efforts aiming at creating comprehensive reference maps of all cell types such as HuBMAP and Fly Cell Atlas. Our lab is part of the ImmGen group and CIFAR Multiscale Human Program. We intensively collaborate with biologists, neuroscientists and medical researchers on cutting-edge problems across fields and disciplines.
Example publications:
- Generative modeling reveals the connection between cellular morphology and gene expression
- Tissue reassembly with generative AI
- Systema: a framework for evaluating genetic perturbation response prediction beyond systematic variation
- immgenT: A Comprehensive Reference of Convergent T-cell States in the Mouse
- Towards Universal Cell Embeddings: Integrating Single-cell RNA-seq Datasets across Species with SATURN
- Annotation of Spatially Resolved Single-cell Data with STELLAR
- MARS: Discovering Novel Cell Types across Heterogeneous Single-Cell Experiments
- Aging Fly Cell Atlas Identifies Exhaustive Aging Features at Cellular Resolution
- High Resolution Single Cell Maps Reveals Distinct Cell Organization and Function Across Different Regions of the Human Intestine
- Fly Cell Atlas: A Single-Cell Transcriptomic Atlas of the Adult Fruit Fly
- Single-Cell Transcriptomes of Developing and Adult Olfactory Receptor Neurons in Drosophila
- Temporal evolution of Single-Cell Transcriptomes of Drosophila Olfactory Projection Neurons
- The Landscape of Microbial Phenotypic Traits and Associated Genes