跳到正文
返回论文列表
cs.CE提交于 已译

RINS: 大型稀疏线性方程组的残差像神经子空间求解器

RINS: Residual-Image Neural Subspace Solvers for Large Sparse Linear Systems

Zhongyan Ouyang · Weixin Liao · Mingquan Feng · Yehui Tang · Junchi Yan

中文摘要

大型稀疏线性方程组来自 PDE 离散化,需要修正子空间,其算子像用于解释当前残差。研究这种残差像视角并提出 Gate-RINS,一种神经子空间求解器,从缓存的残差探测生成多项式修正基,并用轻量级的残差和坐标依赖的点向门进行调制。投影最小二乘更新保持不变,因此神经组件仅选择扩展方向,数值闭包通过 range(AQ_t) 测试它们。还引入混合控制器调度,在同一投影求解器下组合 GRANS 风格图控制器与 Gate-RINS。在六个 PDE 衍生基准任务和两种规模下,Gate-RINS 在同步时钟时间内比 GMRES 和最近的纯图神经基线更快达到固定相对残差阈值,混合调度进一步改善残差轨迹。困难模式诊断和轨迹可视化支持这些增益与算子像子空间更有效对齐当前残差的解释。

关键要点

  1. 01问题:大型稀疏线性方程组需要算子像与当前残差对齐的修正子空间
  2. 02方法:用缓存残差探测生成多项式修正基,通过点向门选择扩展方向,数值闭包通过 range(AQ_t) 测试
  3. 03结果:在六个 PDE 基准任务上,Gate-RINS 比 GMRES 和纯图神经基线更快达到相对残差阈值
  4. 04结果:混合调度(GRANS 图控制器+Gate-RINS)进一步改善残差收敛轨迹
  5. 05局限:困难模式诊断表明增益与算子像子空间对齐残差相关,但未明确指出计算开销对比

解读

尚无解读。

原始英文摘要

arXiv:2610.02217v1 Announce Type: cross Abstract: Large sparse linear systems from PDE discretizations require correction subspaces whose operator images explain the current residual. We study this residual-image viewpoint and propose Gate-RINS, a neural subspace solver that generates polynomial correction bases from cached residual probes and modulates them with a lightweight residual- and coordinate-dependent pointwise gate. The projected least-squares update remains unchanged, so the neural component only chooses the expansion directions while the numerical closure tests them through \(\operatorname{range}(AQ_t)\). We also introduce a hybrid controller schedule that composes GRANS-style graph controllers with Gate-RINS under the same projected solver. Across six PDE-derived benchmark tasks and two scales, Gate-RINS reaches fixed relative-residual thresholds faster in synchronized wall-clock time than GMRES and a recent graph-only neural baseline in most settings, and hybrid schedules further improve the residual trajectory. Difficult-mode diagnostics and trajectory visualizations support the interpretation that these gains are associated with operator-image subspaces that align more effectively with the current residual.

同方向论文 · cs.CE

查看全部 →