Search papers, labs, and topics across Lattice.
This paper introduces Agentic Symbolic Search (ASYS), a framework that leverages prior knowledge and evolutionary search to generate testable differentiable symbolic programs for characterizing partial differential equations (PDEs). By refining mathematical forms through a combination of evolutionary search and gradient-based optimization, ASYS not only recovers known analytical solutions but also constructs new analytical approximations for complex problems. The results demonstrate ASYS's ability to produce interpretable representations for PDEs, marking a significant advancement beyond traditional numerical and neural network approaches.
ASYS reveals a novel way to automate the discovery of analytical forms for PDEs, producing interpretable solutions where none existed before.
Mathematicians understand a PDE solution through mathematical structures rather than tables of computed values. Historically, this has been the product of mathematical analysis, carried out by hand for each problem individually. Neither numerical simulation nor neural networks produce those structures directly. We propose Agentic Symbolic Search (ASYS), a prior-guided framework in which an agent translates PDE theory, public problem constraints, and accumulated search experience into testable differentiable symbolic programs. The mathematical forms are refined under evolutionary search, while their continuous parameters are fit by gradient-based optimization. This makes the search an automated form of inductive-bias injection rather than blind symbolic regression. For problems with known analytical forms, ASYS recovers these forms naturally; for other problems, ASYS constructs analytical approximations which can guide mathematicians toward further analysis. In our experiments, across five problems spanning bounded dynamics, finite-time blow-up, and free-boundary focusing, ASYS produces interpretable representations, including a geometric interface formula for Allen-Cahn 2D dynamics and a nine-parameter contraction law for Keller-Segel chemotactic blow-up, in settings where no closed-form description was previously available. ASYS shows the possibility of a new paradigm for characterizing PDE solutions, beyond handcrafted analytical solutions, mesh-based numerical solutions, and neural network approximations.
NUS