System-Anchored Knee Estimation for Low-Cost Context Window Selection in PDE Forecasting

TL;DR

This paper introduces SAKE, a novel low-cost algorithm for selecting the optimal context window in neural PDE simulators, improving efficiency and performance across multiple PDE benchmarks.

Contribution

The paper formalizes the context-window selection problem and proposes SAKE, a two-stage system-anchored knee estimation method that outperforms existing approaches in efficiency and accuracy.

Findings

SAKE achieves 67.8% Exact and 91.7% Within-1 accuracy.

SAKE reduces search cost by 94.9%.

SAKE outperforms other methods across eight PDEBench families.

Abstract

Autoregressive neural PDE simulators predict the evolution of physical fields one step at a time from a finite history, but low-cost context-window selection for such simulators remains an unformalized problem. Existing approaches to context-window selection in time-series forecasting include exhaustive validation, direct low-cost search, and system-theoretic memory estimation, but they are either expensive, brittle, or not directly aligned with downstream rollout performance. We formalize explicit context-window selection for fixed-window autoregressive neural PDE simulators as an independent low-cost algorithmic problem, and propose \textbf{System-Anchored Knee Estimation (SAKE)}, a two-stage method that first identifies a small structured candidate set from physically interpretable system anchors and then performs knee-aware downstream selection within it. Across all eight PDEBench families evaluated under the shared \(L\in\{1,\dots,16\}\) protocol, SAKE is the strongest overall matched-budget low-cost selector among the evaluated methods, achieving 67.8\% Exact, 91.7\% Within-1, 6.1\% mean regret@knee, and a cost ratio of 0.051 (94.9\% normalized search-cost savings).