AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

MeshTok: Efficient Multi-Scale Tokenization for Scalable PDE Transformers

Yanshun Zhao, Xiaoyu Peng, Jiamin Jiang, Congcong Zhu, Jingrun Chen

ICML 2026 · Architectures and efficiency › Attention and transformers · approach: Block-sparse attention kernels · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add activity-based adaptive refinement (top 25% patches) to coarse tokenization (ablation)MeshTok Ours-Big · Averaged benchmarks, 1-steprelative L2 error: 1.138 -> 0.972Improved
add activity-based refinement (ablation)MeshTok Ours-Big · Averaged benchmarks, 10-step rolloutrelative L2 error: 2.607 -> 2.261Improved
refine all patches (full refinement) instead of partial refinement (ablation)MeshTok Ours-Small · Averaged benchmarks, 1-steprelative L2 error: 1.979 -> 2.014Hurt
refine all patches instead of partial refinement (ablation)MeshTok Ours-Large · Averaged benchmarks, 1-steprelative L2 error: 0.837 -> 0.716Improved
match MACs by widening or deepening a coarse-token baseline (ablation)Uniform coarse ViT-style baseline · Compute-matched, Ours-Bigrelative error: 0.9723 (MeshTok) vs 1.0294 and 1.0652Hurt
full refinement with reduced width or depth to match MACs (ablation)Uniform full-refinement baseline · Compute-matched, Ours-Bigrelative error: 0.9723 (MeshTok) vs 1.0777 and 1.0632Hurt
pretrain on source PDEs before fine-tuning (ablation)MeshTok · Shear flow, Burgers, Reaction-Diffusionrelative L2 error: Shear 3.817 -> 3.334; Burgers 0.324 -> 0.252; React-Diff 11.096 -> 8.106Improved
pretrain on source PDEs before fine-tuning (ablation)MeshTok · Black-Scholes-Barenblattrelative L2 error: 0.127 -> 0.142Hurt
increase model scale from Small to Large with AMR refinement (ablation)MeshTok · Averaged benchmarks, 1-steprelative L2 error: 1.639 -> 0.837Improved
use MeshTok instead of BCATPDE Transformer · PDEBench CNS (0.1,0.01)relative L2 error: 0.254 -> 0.249No clear effect
use MeshTok instead of BCATPDE Transformer · PDEBench SWErelative L2 error: 0.443 -> 0.276Improved
add AMR refinement in 3D (k = 0.375) versus no refinement (ablation)MeshTok · The Well MHD (3D)relative error: 39.647 -> 35.851Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy