Search papers, labs, and topics across Lattice.
This paper introduces AutoPDE, a novel code agent designed to enhance the reliability of numerical solvers for partial differential equations (PDEs) by explicitly representing solver strategies throughout the solving process. By maintaining an independent and inspectable solver strategy, AutoPDE allows for iterative revisions based on numerical evidence, addressing the limitations of traditional LLM-based coding agents that often obscure the underlying decision-making process. Experimental results demonstrate that AutoPDE achieves a pass rate of 54.5% on the PDE Agent Bench, outperforming the strongest baseline by 14.2 percentage points, highlighting its effectiveness in improving solver reliability.
AutoPDE reveals that explicitly representing solver strategies can boost PDE solving reliability by over 14% compared to existing methods.
Numerical solvers for partial differential equations (PDEs) are core computational tools in science and engineering. Building reliable PDE solvers requires not only executable code, but a numerical solver strategy, a set of decisions about discretization, stabilization, solver configuration, and resolution control, that matches the PDE structure. Recent LLM-based coding agents have begun to reduce the programming burden by generating and debugging solver implementations. However, they typically move directly from a PDE problem to solver code, leaving the solver strategy implicit in implementation details. Feedback from a failed solve is therefore routed back to code edits rather than to the underlying strategy, so numerical decisions remain hard to check before code is generated and hard to revise using numerical evidence when it fails. To address this limitation, we propose AutoPDE, a code agent that maintains the solver strategy as an explicitly represented object throughout the solving process: an independent, inspectable object that is built before any code is written and can be revised, using numerical evidence, whenever a solve fails. AutoPDE builds and maintains this object in three stages, all drawing from a library of reusable PDE-solving skills: PDE analysis identifies the equation type and algebraic structure; numerical method selection chooses a numerical method that matches the analysis result and commits to a discretization, stabilization, and linear solver accordingly; and adaptive tuning runs low-cost pilot solves to calibrate resolution and tolerances under the prescribed accuracy and runtime budget. We evaluate AutoPDE on the PDE Agent Bench, where experimental results show that AutoPDE achieves a pass rate of $54.5%$, improving over the strongest baseline by $14.2$ percentage points.
Tsinghua AI