- Cell2Sentence project, fine-tuning LLMs for tasks like cell generation and annotation by encoding scRNA-seq profiles as “cell sentences”. (ICML 2024)
- C2S-Scale framework scaled to 27 billion parameters, achieving state-of-the-art performance for complex multicellular analyses. (bioRxiv 2025 Preprint)
- CINEMA-OT applies causal inference with optimal transport to disentangle true treatment effects from confounders in single-cell perturbation experiments. (Nature Methods 2023)
- 'Intelligence at the Edge of Chaos' pinpointed a “sweet spot” of data complexity that maximizes downstream predictive and reasoning abilities. (ICLR 2025)
- MAGIC leverages Markov affinity-based graph diffusion to impute missing transcripts in single-cell RNA-seq data.
Research Experience
- Presented 'Intelligence at the Edge of Chaos' at ICLR 2025.
- Released C2S-Scale preprint.
- Gave a talk at the Broad Institute.
Education
No specific educational background information provided
Background
Research Interests: Application of artificial intelligence and large-scale foundation models in biomedicine; Specializes in rigorous mathematics, state-of-the-art ML, and rich genomic and clinical data.