# Mohammad Asadi

> PhD student in Electrical Engineering at Stanford University, advised by Prof. Euan Ashley, Prof. Fei-Fei Li, and Prof. Ehsan Adeli. Works on multimodal and agentic AI, AI evaluation, and trustworthy AI, with current applications in medicine. Amazon AI PhD Fellow and Stanford HAI Graduate Fellow.

Mohammad Asadi is a PhD student in the Department of Electrical Engineering at Stanford University, advised by Prof. Euan Ashley and co-advised by Prof. Fei-Fei Li and Prof. Ehsan Adeli (Ashley Lab and the Stanford Translational AI (STAI) Lab). His research has two complementary threads: developing multimodal, agentic AI systems that reason jointly over heterogeneous data, and building evaluations that expose where frontier multimodal models fail. His current work applies these methods in medicine. He is the first author of MIRAGE (a study of "mirage reasoning," where vision-language models confidently describe inputs they were never shown), which was covered by Fortune, Live Science, and Futurism, and co-first author of MARCUS, an agentic multimodal system for cardiac diagnosis. He has previously worked on human-motion generation at Samsung and interpretable machine learning for education at EPFL, and earned a BSc in Electrical Engineering at Sharif University of Technology.

## Key pages
- [Home / About](https://masadi.ai/): bio, news, press, selected publications.
- [MIRAGE](https://masadi.ai/mirage/): project page for MIRAGE, with key results and press.
- [MARCUS](https://masadi.ai/marcus/): project page for MARCUS, with key results.
- [Publications](https://masadi.ai/publications/): full list.
- [In The Press](https://masadi.ai/press/): media coverage of his research.
- [CV](https://masadi.ai/cv/): education, experience, awards, patents.

## Selected publications
- Align-RAG: Alignment Is All You Need for TSFM In-Context Learning (first author, 2026, collaboration with Amazon): https://arxiv.org/abs/2608.05571
- MIRAGE: The Illusion of Visual Understanding (first author, 2026): https://arxiv.org/abs/2603.21687
- MARCUS: An Agentic, Multimodal Vision-Language Model for Cardiac Diagnosis and Management (co-first author, 2026): https://arxiv.org/abs/2603.22179
- Deterministic Hallucination Detection in Medical VQA via Confidence-Evidence Bayesian Gain (first author, accepted to MICCAI 2026): https://arxiv.org/abs/2603.21693
- EchoGraph-annotated ECHO-NOTE2NUM examples (dataset, PhysioNet 2025, DOI 10.13026/hb5q-9532): https://physionet.org/content/echograph-note2num-annotations/1.0.0/
- OpenMHC: Accelerating the Science of Wearable Foundation Models (co-author, 2026): https://arxiv.org/abs/2607.16235
- NeuroQA: A Large-Scale Image-Grounded Benchmark for 3D Brain MRI Understanding (co-author, 2026): https://arxiv.org/abs/2605.20525
- Ripple: Concept-Based Interpretation for Raw Time Series Models in Education (first author, AAAI 2023): https://arxiv.org/abs/2212.01133

## Facts
- Affiliation: Department of Electrical Engineering, Stanford University; Ashley Lab; Stanford Translational AI (STAI) Lab.
- Advisors: Prof. Euan Ashley (advisor); Prof. Fei-Fei Li and Prof. Ehsan Adeli (co-advisors).
- Awards: Amazon AI PhD Fellowship; Stanford HAI Graduate Fellowship; "Smartest AI Agent" Prize at TreeHacks 2024.
- Press: MIRAGE covered by Fortune, Live Science (interview), and Futurism (interview), with international coverage including GIGAZINE (Japan), Al Khaleej (UAE), Tencent News (China), AI Sparkup (Korea), The Decoder (Germany), Génération-NT and Acuité (France), and BAAI (China). Full list at https://masadi.ai/press/.
- Expert commentary: quoted by Live Science (September 2026) as an independent expert on Pathology-CoT, a Nature Biomedical Engineering study of AI for pathology; not an author of that study.
- Uptake: MIRAGE is cited by independent groups, including "Mirage Probes: How Vision Models Fake Visual Understanding" (arXiv 2606.13870). Mirage Score, Phantom-0, and B-Clean are openly released.
- MIRAGE is also hosted as a Stanford Graduate School of Business working paper: https://www.gsb.stanford.edu/faculty-research/working-papers/mirage-illusion-visual-understanding
- Citations: approximately 125 total (Google Scholar, July 2026).
- Contact: masadi@stanford.edu · Google Scholar (cRuOHB0AAAAJ) · GitHub (masadi-99) · LinkedIn (mohammad-asadi-9a7b6b191).