Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Transolver is a Linear Transformer: Revisiting Physics-Attention through the Lens of Linear Attention

About

Recent advances in Transformer-based Neural Operators have enabled significant progress in data-driven solvers for Partial Differential Equations (PDEs). Most current research has focused on reducing the quadratic complexity of attention to address the resulting low training and inference efficiency. Among these works, Transolver stands out as a representative method that introduces Physics-Attention to reduce computational costs. Physics-Attention projects grid points into slices for slice attention, then maps them back through deslicing. However, we observe that Physics-Attention can be reformulated as a special case of linear attention, and that the slice attention may even hurt the model performance. Based on these observations, we argue that its effectiveness primarily arises from the slice and deslice operations rather than interactions between slices. Building on this insight, we propose a two-step transformation to redesign Physics-Attention into a canonical linear attention, which we call Linear Attention Neural Operator (LinearNO). Our method achieves state-of-the-art performance on six standard PDE benchmarks, while reducing the number of parameters by an average of 40.0% and computational cost by 36.2%. Additionally, it delivers superior performance on two challenging, industrial-level datasets: AirfRANS and Shape-Net Car.

Wenjie Hu, Sidun Liu, Peng Qiao, Zhenglun Sun, Yong Dou• 2025

Related benchmarks

TaskDatasetResultRank
PDE solvingNavier-Stokes Regular Grid (test)
Relative L2 Error0.0699
41
PDE solvingDarcy Regular Grid (test)
Relative L2 Error0.005
41
PDE solvingPipe Structured Mesh (test)
Relative L2 Error0.0024
38
PDE solvingAirfoil Structured Mesh (test)
Relative L2 Error0.0049
38
PDE solvingPlasticity Structured Mesh (test)
Relative L2 Error0.0011
23
Aerodynamic SimulationAirfRANS (test)
Volume MSE0.0011
22
PDE solvingElasticity Point Cloud (test)
Relative L2 Error0.5
20
Physical Property PredictionShapeNet Car (test)
Volume0.0194
9
Showing 8 of 8 rows

Other info

Follow for update