The lab

Daniel Levine

Postdoctoral Associate · Yale–Boehringer Ingelheim Biomedical Data Science Fellow

Daniel joined as a postdoc in 2023 and led the original Cell2Sentence paper, which first taught large language models to read and write single cells. Daniel came to machine learning from pure mathematics (PhD, Penn State, on moduli spaces of vector bundles) and now works on AI systems that reason about genetics and on biologically inspired learning rules for continual learning.

Syed Asad Rizvi

PhD Student · Computer Science

Google PhD Fellowship (2025)

Syed is lead author of C2S-Scale, the 27-billion-parameter model of single-cell biology built with Google. The work extends language models from single cells to tissues and patients: prompt-conditioned cell embeddings, multi-cell tokenization, reinforcement learning for perturbation prediction, and a language for DNA methylation. Syed holds a BS from the University of Houston.

Sizhuang He

PhD Student · Computer Science

Sizhuang develops generative models for data with unusual structure: flow matching with causal language models (CaLMFlow), diffusion over permutations (ICML 2026), and edit-based post-training for biological sequences. Sizhuang also builds multi-agent systems that harmonize public epigenetic data for aging research. Sizhuang studied mathematics at the University of Michigan.

Shiyang (Harry) Zhang

PhD Student · Computer Science

Harry studies where reasoning in language models comes from. The paper "Intelligence at the Edge of Chaos" (ICLR 2025) showed that models trained on data of the right complexity become better reasoners. Harry now works on multi-agent systems (MoRSE, NeurIPS 2026) and on models that reason about chemistry mechanistically. Harry holds an MS from Columbia and a BSc in Statistics from the University of British Columbia.

Aakash Patel

PhD Student · Computer Science

Salovey–Moret Data Science Fellow (2026)

Aakash builds foundation models and agents for neuroscience and genomics. Aakash is the student lead of the Wu Tsai Institute Foundation Models Grand Challenge and first author of The AI Neuroscientist, an agent that runs neuroimaging analyses through conversation. Aakash was named to the inaugural cohort of Yale’s Salovey–Moret Data Science Fellows. Aakash holds dual master’s degrees in CS and Mathematics from the University of Michigan.

Yangtian Zhang

PhD Student · Computer Science

Yangtian, co-advised with Rex Ying, works on the foundations of generative modeling for sequences. Yangtian led CaDDi (NeurIPS 2025), which unifies causal language models with discrete diffusion, and a variational approach to insertion-based generation with collaborators at Google DeepMind and UCL (ICML 2026).

Peiwen Li

PhD Student · Computer Science

Peiwen joined in 2025, co-advised with Rex Ying, and designs multi-agent LLM systems that divide scientific and reasoning tasks among specialized roles. Peiwen is first author of MoRSE (NeurIPS 2026), a mixture of role-and-subtask experts for task-oriented agent teams, with broader interests in graph machine learning and causal discovery.

Daanish Qureshi

PhD Rotation Student · Computational Biology & Biomedical Informatics

Daanish is a first-year CBB PhD student rotating in the lab in fall 2026, working on multi-agent brain simulation and on when feedback-driven iteration beats stored knowledge.

Ivan Vrkic

Postgraduate Researcher

Ivan works on scaling Cell2Sentence models and on language models for scientific inverse design, including SciDesignBench. Ivan studied at EPFL and TU Wien and previously built control software at CERN.

William Wang

Undergraduate Researcher · Yale College

William works with Syed Rizvi on extensions of Cell2Sentence and on the lab’s few-shot optimality project.