PDE-JEPA: Predictive Representation Learning of Latent Dynamics Modeling for Parametric PDEs

arXiv 2026 23.5 method

TLDR

PDE-JEPA pretrains encoders with masked latent prediction and geometry-aligned predictors to model parametric PDE dynamics, outperforming state-of-the-art on nine benchmarks.

Reasoning

The paper introduces a predictive representation learning framework with a geometry projector and physics-structured predictor, showing strong empirical gains across benchmarks. However, its scope is limited to PDE dynamics and it does not engage with broader world model or interactive agent settings.

Read-first score

Read-first score 23.5, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 9.

Recency 5%

100

Uses a gentle age decay so recent papers surface without erasing older foundations. 2026

Methodology quality 15%

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=benchmark,experiment,metric

Reproducibility 15%

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=False; dataset=False; markers=code,github

Topical relevance 25%

12.9

Uses existing LLM keyword relevance scores normalized to 0-100. world model,world simulator,generative world model,interactive world model,video world model,world dynamics prediction,model-based reinforcement learning world model

Graph prestige 15%

5.7

PageRank-style prestige over the local citation graph built from available OpenAlex references. local_pagerank=0.0016

Citation impact 15%

Uses OpenAlex-shaped citation metadata as a bibliometric attention signal, separate from paper quality. cited_by_count=0

Citation velocity 10%

Citation velocity estimates citations per publication-year to reduce old-paper bias. velocity=0.00

Field roles

Frontier

Rank sensitivity

Stability: volatile; rank range: 147.

Keyword Scores

world dynamics prediction

world model

world simulator

generative world model

interactive world model

video world model

model-based reinforcement learning world model

Resources

Paper