arXiv cs.AI· Siqi Gu, Gaozhen Ma, Liheng Huang, Weiqi He, Jiawen Luo, Jiaming Wu, Lixin Zheng·· 13 小时前AI 评分34
DPR：解耦剪枝与重建，在紧凑核心上训练的双层稀疏化方法
Decoupled Prune-and-Reconstruct: Training on a Compact Core with Double-Layer Sparsification
AI 导读
DPR 提出“先剪枝提效、再重建补精度”的解耦框架，用同一紧凑节点集同时约束编码器训练图与推理锚点，由稀疏结构感知解码器重建被剪节点。30% 预算模型在单张 6 GB GPU 上以 1,450 MiB 峰值内存训练，恢复稠密重建参考的 96.9%，比同 GPU 的 DSpar 高 1.36 pp 且峰值训练内存低 50.1%。
正文 · 原文
Abstract:On million-node graphs, the memory wall arrives before the compute wall: a GNN that never fits in VRAM cannot deliver the all-node inference its accuracy promises. The standard remedy, pruning, has always been an accuracy tax - paid once and never refunded. DPR refuses this framing. Its central insight is that the mask that shrinks training can also scaffold recovery: one compact node set bounds both computations of graph learning, serving as the encoder's training graph and, at inference, as the anchor from which a sparse structure-aware decoder rebuilds every pruned node (double-layer sparsification). The order is inverted - prune for efficiency, then reconstruct for accuracy - and the tax is repaid: the deployed 30%-budget model trains in 1,450 MiB of peak memory on one 6 GB GPU and recovers 96.9% of its dense-reconstruction reference, beating a same-GPU rerun of the edge-sparsification baseline DSpar by +1.36 pp at 50.1% lower peak training memory; against the same-protocol full-batch no-prune encoder the deployed point pays a measured 2.64 pp accuracy tax for a 2.8x smaller training footprint. Pruning without reconstruction leaves the encoder at 0.378 within the same 40-epoch window (Fig. 4), and all-node serving runs 46x faster than sampling-based GraphSAGE, giving up approx. 5.4 pp of test accuracy. The framework scales to a 2.45M-node production graph at 6,285 MiB of peak training memory; extended variants (long-range context, external text features, a CAN-bus application) are reported as extensions and do not compete with the deployed method, and the 41.8%-budget M4 point, scored on a coarsened super-node graph with its own split, is reported separately rather than compared with the original-node baselines. DPR thus turns aggressive node pruning from an accuracy tax into a safe training accelerator for VRAM-limited all-node inference.
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.34140 [cs.LG] |
| (or arXiv:2610.34140v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.34140 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Siqi Gu [view email]
[v1]
Sun, 25 Oct 2026 08:20:30 UTC (490 KB)
来源：arXiv cs.AI · arxiv.org