LTesan/MPNN-UnderReach
MPNN-UnderReach explores the under-reaching phenomenon in message-passing neural PDE solvers. We propose physics-guided lower bounds on the number of message passing iterations. These bounds link graph information propagation to the CFL condition and PDE type. Official code for Tesan L. & Iparraguirre M.M. (2025)
What's novel
MPNN-UnderReach explores the under-reaching phenomenon in message-passing neural PDE solvers. We propose physics-guided lower bounds on the number of message passing iterations. These bounds link graph information propagation to the CFL condition and PDE type. Official code for Tesan L. & Iparraguirre M.M. (2025)
Score Breakdown
Signal breakdown
Innovation
Craft
Traction
Scope
Evidence
Commits
18
Contributors
1
Files
690
Active weeks
2
Repository
Language
Python
Stars
2
Forks
0
License
GPL-3.0