Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in Coakley et alK. L. Coakley, T. Snelleman, H. Hoos, and O. E. Gundersen, "The embrace of open science: An analysis of a decade of AI research and 56 800 conference papers," Under Review, 2026..
Boundary-Value PDEs Meet Higher-Order Differential Topology-aware GNNs
Authors: Yunfeng Liao, Yangxin Wu, Xiucheng Li
NeurIPS 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experiments show that our proposed method outperforms the existing neural operators by large margins on BVPs in electromagnetism. Our code is available at https://github. com/Supradax/Higher-Order-Differential-Topology-aware-GNN. |
| Researcher Affiliation | Academia | Yunfeng Liao Yangxin Wu Xiucheng Li School of Computer Science and Technology, Harbin Institute of Technology (Shenzhen) EMAIL |
| Pseudocode | No | The paper describes methodologies and mathematical formulations but does not include explicit pseudocode or algorithm blocks. |
| Open Source Code | Yes | Our code is available at https://github. com/Supradax/Higher-Order-Differential-Topology-aware-GNN. |
| Open Datasets | No | The data on electromagnetic static fields will be uploaded to the repository. Their details can be found in Appendix E. |
| Dataset Splits | Yes | Datasets are partitioned into training, validation, and test sets with the ratio 8:1:1. |
| Hardware Specification | Yes | The hardware configuration consists of an Ubuntu 20.04 LTS operating system platform powered by four NVIDIA RTX A6000 Graphics Processing Units with 48GB memory. |
| Software Dependencies | Yes | We use the Py Troch framework version 1.13 alongside Python 3.12 interpreter. |
| Experiment Setup | Yes | Model optimization involved selection of the initial learning rate parameter γ for the Adam optimizer, which was systematically chosen from the discrete value set {i 10 j : i {1, 5}, j {1, 2, 3, 4, 5}} to achieve optimal performance metrics. Each model in the main result is trained until the validation loss converges within 1,000 epochs. |