Skip to main content
QUICK REVIEW

[论文解读] Prune, Don't Rebuild: Efficiently Tuning $α$-Reachable Graphs for Nearest Neighbor Search

Tian Zhang, Ashwin Padaki|arXiv (Cornell University)|Feb 8, 2026
Advanced Image and Video Retrieval Techniques被引用 0
一句话总结

RP-Tuning 提供了一种事后剪枝例程,用以在不重构索引的情况下调整 DiskANN 的 α-可达性,并带有理论保证,在不同数据集上实现高达 43× 的调优加速。

ABSTRACT

Vector similarity search is an essential primitive in modern AI and ML applications. Most vector databases adopt graph-based approximate nearest neighbor (ANN) search algorithms, such as DiskANN (Subramanya et al., 2019), which have demonstrated state-of-the-art empirical performance. DiskANN's graph construction is governed by a reachability parameter $α$, which gives a trade-off between construction time, query time, and accuracy. However, adaptively tuning this trade-off typically requires rebuilding the index for different $α$ values, which is prohibitive at scale. In this work, we propose RP-Tuning, an efficient post-hoc routine, based on DiskANN's pruning step, to adjust the $α$ parameter without reconstructing the full index. Within the $α$-reachability framework of prior theoretical works (Indyk and Xu, 2023; Gollapudi et al., 2025), we prove that pruning an initially $α$-reachable graph with RP-Tuning preserves worst-case reachability guarantees in general metrics and improved guarantees in Euclidean metrics. Empirically, we show that RP-Tuning accelerates DiskANN tuning on four public datasets by up to $43 imes$ with negligible overhead.

研究动机与目标

  • 理解并优化基于 DiskANN 的 ANN 搜索中准确性、延迟和索引大小之间权衡的动机。
  • 开发一种在不进行完整索引重建的情况下调整可达性参数 α 的方法。
  • 为修剪后的 worst-case 性能提供理论保证。
  • 在公开数据集上实证展示速度和准确性提升。

提出的方法

  • 提出 RP-Tuning,一种从 RobustPrune 推导的事后剪枝例程,用于 DiskANN 构建。
  • 以 α1 可达的基图出发,剪裁至目标 α2 < α1,且保留最坏情况保证。
  • 在一般度量下证明剪裁后图的最坏情况可达性界限,在欧几里得度量下给出改进界限(包括排序/未排序变体)。
  • 利用现有的 α 可达性理论分析剪裁后的构建、查询时间和近似性。
  • 在四个公开数据集(SIFT-1M、GIST-1M、Deep-1M、MSSPACEV-1M)上通过 100-recall@100 和不同的 beam size L,实证比较 RP-Tuning 与重建的差异。
Figure 3 : Recall-QPS trade-off frontiers achieved by a base DiskANN graph of $\alpha_{1}=1.2$ then pruned graphs from the base graph via RP-Tuning with $\alpha_{2}=1.1,1.05,1.01$ , and the same base DiskANN graph of $\alpha=1.2$ along with rebuilt DiskANN graphs of $\alpha=1.1,1.05,1.01$ .
Figure 3 : Recall-QPS trade-off frontiers achieved by a base DiskANN graph of $\alpha_{1}=1.2$ then pruned graphs from the base graph via RP-Tuning with $\alpha_{2}=1.1,1.05,1.01$ , and the same base DiskANN graph of $\alpha=1.2$ along with rebuilt DiskANN graphs of $\alpha=1.1,1.05,1.01$ .

实验结果

研究问题

  • RQ1RP-Tuning 能否将 α1 可达的 DiskANN 图裁剪到更小的 α2,同时保持可预测的 worst-case 保证?
  • RQ2排序与未排序的 RobustPrune 变体在一般和欧几里得空间中的最坏情况可达性有何影响?
  • RQ3相较于对不同 α 值进行索引重建,RP-Tuning 的实际加速和准确性影响是什么?
  • RQ4在不同数据集上裁剪是否能保持或改善实际的召回-延迟权衡?

主要发现

DatasetT_rebuild(alpha=1.01) (s)T_prune(alpha=1.01) (s)Speedup(alpha=1.01)T_rebuild(alpha=1.05) (s)T_prune(alpha=1.05) (s)Speedup(alpha=1.05)T_rebuild(alpha=1.10) (s)T_prune(alpha=1.10) (s)Speedup(alpha=1.10)Total_T_rebuild (s)Total_T_prune (s)Total_Speedup
SIFT-1M5,72840914×6,22042115×6,72848214×18,6761,31214×
DEEP1M5,53637615×5,91149912×6,74555512×18,1921,43013×
GIST-1M12,07329441×14,66536740×21,31246246×48,0501,12343×
MSSPACEV1M7,57349215×9,02856016×11,57049523×28,1711,54718×
  • RP-Tuning 在四个公开数据集上将 DiskANN 调优加速至多 43×。
  • 通过 RP-Tuning 剪裁的图在相同 α 下,与重建索引相比,在召回-查询每秒(QPS)方面呈现有利的权衡。
  • 理论结果给出剪裁后图的最坏情况可达性界限,排序剪裁在欧几里得场景提供更好的保证。
  • 剪裁从高度连通的基图(α1)开始,产生可配置的更稀疏图(α2),适用于资源受限的部署。
  • 在经验上,尽管存在潜在的最坏情况保证,裁剪后的索引在 QPS-召回性能方面通常优于具有相同 α 的重建索引。
  • 剪裁保留了基图的结构完整性优势,重建无法轻易复制。
Figure 4 : Recall-QPS trade-off frontiers achieved by a base DiskANN graph of $\alpha_{1}=1.2$ (blue curves with circles) and pruned graphs (curves with squares) from the base graph of $\alpha_{2}=1.1,1.05,1.01$ . Average degrees (Deg) of individual graphs are also included.
Figure 4 : Recall-QPS trade-off frontiers achieved by a base DiskANN graph of $\alpha_{1}=1.2$ (blue curves with circles) and pruned graphs (curves with squares) from the base graph of $\alpha_{2}=1.1,1.05,1.01$ . Average degrees (Deg) of individual graphs are also included.

更好的研究,从现在开始

从论文设计到论文写作,大幅缩短您的研究时间。

无需绑定信用卡

本解读由 AI 生成,并经人工编辑审核。