Announcement Detail
Thursday, June 18, 2026
1:00 PM EDT
Join via Zoom: https://us06web.zoom.us/j/82464478256?pwd=ZMkJVFdjMJzadgnVWFPqsdUSs4qTaY.1
Student Chapter Seminar Series
GRAFT-ATHENA: Self-Improving Agentic Teams for Autonomous Discovery and Evolutionary Numerical Algorithms
Speaker
Juan Diego Toscano, Brown University
Abstract:
Scientific discovery can be modeled as a sequence of probabilistic decisions that map physical problems to numerical solutions. Recent agentic AI systems automate individual scientific tasks by orchestrating LLM-driven planners, solvers, and evaluators. Each method is a combination of methodological actions, with many viable combinations for any given problem and structural dependencies between choices. However, existing frameworks treat each problem in isolation, with no shared substrate to accumulate methodological experience across domains. Here we show that GRAFT-ATHENA, a self-improving agentic framework, learns from past problems and autonomously expands its own action space across diverse domains. GRAFT (Graph Reduction to Adaptive Factored Trees) projects combinatorial decision spaces into factored probabilistic trees in which each method is a single path, taking the parameter footprint from exponential to linear. In the lineage of classical Bayesian networks, the factorization is an $I$-map of the policy, and the resulting paths embed as unique fingerprints in a metric space whose closeness lets each new problem learn from similar past ones. On canonical physics-informed machine learning (PIML) benchmarks, GRAFT-ATHENA reaches near-machine-precision accuracy, surpassing human and prior agentic baselines, and on production solvers, it tackles complex engineering problems such as reconstructing Mach-10 flow over the Apollo Command Module from a 1968 report and recovering shear-thinning blood-cell rheology. Notably, the system grows its own knowledge substrate, autonomously proposing regularization constraints for ill-posed inverse problems and discovering new numerical methods such as a spectral PINN with exponential convergence. These results provide a foundation for autonomous laboratories that grow more capable with every problem they solve.
Preprint: https://arxiv.org/abs/2605.
Bio:
Juan Diego Toscano is a PhD student in Applied Mathematics at Brown University, advised by Prof. George Em Karniadakis. He received his BS in Mechanical Engineering from Universidad de las Fuerzas Armadas, ESPE (Ecuador). His research focuses on developing reliable and stable machine learning methods to study complex physical systems that resist traditional analysis. To this end, he designs new architectures and optimization methods for scientific machine learning (SciML), with applications to chemical reactors, turbulent thermal convection, and biofluids. More recently, he started exploring agentic systems for scientific discovery, in which teams of agents reason across SciML, classical solvers, and particle-based simulations to propose numerical methods and surface analytical insights, such as well-posedness results and exact solutions.