Research at Meridian AI
Research is the core activity of the institute. Every faculty member at Meridian is primarily a researcher who teaches; no faculty are teaching-only. Every enrolled graduate student participates in funded research from their first semester.
Research Philosophy
Meridian's research is characterized by three commitments that are unusual in combination:
Technical rigor — We build things, run experiments, prove theorems. We do not publish position papers dressed as empirical research. Our empirical claims are supported by careful experimental design and, where possible, by releasing the code and data that would allow independent replication.
Societal grounding — Technical research at Meridian is expected to articulate its relevance to real AI systems and their effects on real people. This does not mean every paper must have a policy section; it means that researchers are expected to know why their work matters beyond citation counts.
Interdisciplinary integration — The most important work in AI requires perspectives from computer science, mathematics, philosophy, economics, law, and cognitive science simultaneously. We design research programs, not just research projects, to ensure these perspectives are in genuine dialogue.
Active Research Projects
Project Lighthouse — Multilingual information retrieval. PI: Dr. Elena Marchetti. Building and evaluating AI-powered search systems that work equitably across high- and low-resource languages. Current focus: ML-IR Bench 2026 and integration with the Scolta search platform.
AlignBench — Reproducible LLM alignment evaluation. PIs: Dr. Kwame Asante, Dr. James Okafor. Developing open evaluation infrastructure for measuring how well language models reflect stated values and follow complex instructions. 18 models currently tracked.
Threshold Robotics Lab — Safe learning for physical systems. PI: Dr. Priya Chakraborty. Research on world models, safe exploration in RL, and sim-to-real transfer for robotic manipulation and navigation.
MEAF (Municipal Ethics and AI Frameworks) — Governance tools for public-sector AI. PI: Dr. Chioma Adeyemi. Developing risk frameworks, audit protocols, and participatory processes for municipal AI adoption. Current deployment: City of Estero Bay partnership.
Open LLM Reproducibility Initiative — Making language model research reproducible. PI: Dr. Amara Osei-Mensah. Maintains open replications of key LLM training experiments and documents the tacit knowledge required to reproduce published results.
Clinical AI Research Group — AI in healthcare. PI: Dr. Fatima Al-Rashid. Applied research on clinical decision support, radiology AI, and health equity, conducted in partnership with Archipelago Health Systems.
Research Publications
Meridian AI faculty and students publish primarily in top-tier ML venues (NeurIPS, ICML, ICLR, ACL, EMNLP, CVPR) and in interdisciplinary venues including Nature Machine Intelligence, AI & Society, and FAccT. All publications are available as preprints through arXiv or the Meridian research portal before or upon submission.
Joining Research at Meridian
Prospective researchers interested in joining Meridian should apply through the graduate admissions process. Faculty do not accept unsolicited research collaborators or unmatriculated research students.
Postdoctoral positions are advertised as they become available on the Meridian jobs portal. We encourage postdoctoral applicants to identify a primary faculty supervisor before applying.