Research Focus
Marlowe enables frontier-scale AI research across Stanford, from 80-billion-parameter brain models to genomic foundation models to autonomous driving systems. Meet the researchers pushing the boundaries of what's possible with GPU computing.
Dan Yamins
Associate Professor of Psychology and Computer Science
School of Humanities & Sciences
Counterfactual World Modeling: Training Brain-Scale Neural Networks
Validated 80B-parameter brain-inspired neural network training on Marlowe in a February 2026 proof-of-concept. Now training PSI2-30B — a 30-billion-parameter counterfactual world model that learns to predict how the physical world changes in response to actions — in a dedicated 30-day campaign across 24 nodes, the first hero run on Marlowe.
Read full story →
Andreas Tolias
Professor of Ophthalmology and of Electrical Engineering
School of Medicine
The Enigma Project: First Foundation Model and Digital Twin of the Brain
Trained the first foundation model of mammalian visual cortex on Marlowe — a 2B-parameter multimodal transformer on recordings from 3 million neurons across 330 mice, establishing the first-ever scaling laws for neuroscience. Now scaling to build a digital twin of the primate brain with up to 1 trillion tokens of neural data across 128 GPUs. One of Marlowe's earliest and most active research groups.
Read full story →
Jure Leskovec
Professor of Computer Science
School of Engineering
AI Virtual Cell: Genomic Foundation Models at Frontier Scale
Building the molecular foundation of the AI Virtual Cell — novel architectures designed for biological sequences (DNA, RNA, proteins) with biologically informed inductive biases, not adaptations of existing language models. Currently at 770M parameters with demonstrated scaling laws, targeting 3B-15B for open-source release comparable to ESM-3 (Science) and EVO-2 (Nature).
Ruijiang Li
Associate Professor of Radiation Oncology
School of Medicine
Virtual Cell World Models for Cancer Biology
Published a vision-language foundation model in Nature (January 2025) for cancer diagnosis and predicting therapeutic response. Now building a generative world model for virtual cells on Marlowe — multi-scale AI that simulates from molecular interactions to tumor microenvironment evolution — integrating histopathology images, clinical notes, spatial transcriptomics, and over 400 million medical images at billion-parameter scale.
Curtis Langlotz
Professor of Radiology and of Biomedical Data Science
School of Medicine
CheXagent: Vision-Language Foundation Models for Radiology
Building CheXagent, a vision-language foundation model for chest X-ray interpretation trained on 8.5 million samples across 35 clinical tasks. In clinical studies, CheXagent reduced radiology report writing time by 36%. Now extending the model on Marlowe to incorporate 3D volumetric CT knowledge into X-ray interpretation, bridging 2D and 3D medical imaging through cross-dimensional representation learning.
Gordon Wetzstein
Associate Professor of Electrical Engineering
School of Engineering
Video World Models for 3D Scene Understanding and Generation
Fine-tuning billion-parameter video diffusion models on Marlowe for 3D-aware scene understanding and generation. Two active projects: real-time automotive novel-view synthesis using LiDAR-conditioned video diffusion for driving simulation, and an in-the-wild 3D foundation model that jointly predicts geometry and novel views from casual photos. Building on Streetscapes (SIGGRAPH 2024) and two CVPR 2025 papers on world-consistent video diffusion and feed-forward 3D estimation.
Brian Hie
Assistant Professor of Chemical Engineering
School of Engineering
Beyond Evo 2: Next-Generation DNA Language Models
Creators of Evo 2, the state-of-the-art DNA language model published in Nature, now developing next-generation bidirectional DNA models on Marlowe to address limitations in protein structure prediction and fitness tasks that scale alone cannot solve. Training comparative models at 150M parameters before scaling to 1B, using the Savanna framework proven across models up to 40B parameters.
Iro Armeni
Assistant Professor of Civil and Environmental Engineering
School of Engineering
4D Scene Understanding: AI for Dynamic Real-World Environments
Two active projects in 4D scene understanding on Marlowe. ReScene4D introduces temporally consistent instance segmentation from sparse 3D scans — a novel task for construction digital twins and embodied AI. A companion project develops end-to-end 4D reconstruction and camera localization from monocular video using DiT architectures for autonomous navigation and AR/VR.
James Zou
Associate Professor of Biomedical Data Science and, by courtesy, of Computer Science and of Electrical Engineering
School of Medicine
AI Agents for Biomedical Discovery: Self-Improving LLMs with Scientific Tools
Developing algorithms that enable large language models to self-improve and learn to use scientific tools like AlphaFold and biomedical databases to build deeper expertise. Fine-tuning and evaluating LLMs (7B-70B parameters) for high-impact applications in healthcare, biology, chemistry, and medicine.
Tengyu Ma
Assistant Professor of Computer Science
School of Engineering
Parallel Chain-of-Thought: Reducing LLM Reasoning Latency via RL
Teaching reasoning LLMs to parallelize their long chains of thought using reinforcement learning — spawning parallel workers that investigate different approaches simultaneously, dramatically reducing inference latency while preserving accuracy on challenging math competitions like AIME. Submitted to ICML 2026, with 1.8x strong scaling demonstrated across 3 Marlowe nodes.
Thierry Tambe
Assistant Professor of Electrical Engineering and, by courtesy, of Computer Science
School of Engineering
Efficient AI Computing: From Model Compression to Edge-Deployable Video Generation
Extending BlockDialect (ICML 2025), their fine-grained mixed-format quantization method, from LLMs to video diffusion transformers — enabling real-time video generation on edge devices by quantizing models like Open-Sora 2.0. Also developing compact visual encoders using 2D Gaussian splatting for vision-language models. One of Marlowe's most active research groups.
Jeffrey Glenn
Joseph D. Grant Professor and Professor of Microbiology and Immunology
School of Medicine
Physics-Guided AlphaFold3: Reinforcement Learning for Drug Discovery
Training AlphaFold3 with reinforcement learning and physics-based Rosetta scoring on Marlowe to produce physically realistic molecular structures for drug discovery. On the CASP16 benchmark, RL training increased the proportion of high-quality structures from 13% to 62% and achieved state-of-the-art binding affinity among deep learning approaches. Now scaling from a 100M-parameter proof-of-concept to the full 400M-parameter AlphaFold3 model across 8 nodes.
Kay Giesecke
Professor of Management Science and Engineering
School of Engineering
Time Machine: Time-Aware Pretrained LLMs for Finance and Economics
Pioneering time-aware LLMs that eliminate look-ahead bias — a fundamental flaw making current models unreliable for finance, economics, and policy analysis. Pretraining novel temporal architectures from 7B to 32B parameters on curated time-ordered datasets with dynamic data masking to prevent information leakage. Aiming to release open-source models and benchmarks as the foundational standard for time-aware AI in finance and economics.
Azalia Mirhoseini
Assistant Professor of Computer Science
School of Engineering
Energy-Efficient AI: From Low-Precision RL to Intelligent Model Orchestration
Training small language model orchestrators on Marlowe that outperform GPT-5 on Humanity's Last Exam at 30% of the computational cost, building on ToolOrchestra. Also pioneering FP8/FP4 precision training for reinforcement learning — enabling 2x larger RL models in the same memory budget — and developing hierarchical text generation beyond token-by-token prediction with Chris Ré. Scaling orchestrators to 32B parameters with 200K+ training trajectories.