Why Are Learned Indexes So Effective but Sometimes Ineffective?
Summary: The paper gives a theoretical account of PGM-Index effectiveness and pins practical slowdown to memory-bound queries and expensive internal search. Introduces PGM++—a mix-of-search-strategies with cost-model auto-tuning—achieving much faster lookups at similar space. (summarized by gpt-5-mini on Feb 09 2026)
Incoming Non-self Citations Over Time
Authors
- 1. Qiyu Liu
- 2. Siyuan Han
- 3. Yanlin Qi
- 4. Jingshu Peng
- 5. Jin Li
- 6. Longlong Lin
- 7. Lei Chen
Incoming Citations (Sorted by Pagerank)
Showing 2 of 2 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 10,172 | HIRE: A Hybrid Learned Index for Robust and Efficient Performance under Mixed Workloads | 2026 | SIGMOD | 4.1945683e-05 |
| 10,698 | Not Small Enough? SegPQ: A Learned Approach to Compress Product Quantization Codebooks | 2025 | VLDB | 4.1945683e-05 |
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 20 of 20 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
Previous
Page 1 / 1
Next
Semantically Similar Papers
| Overall Rank | Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 9,346 | Can Learned Indexes be Built Efficiently? A Deep Dive into Sampling Trade-offs | 2024 | SIGMOD | 4.3532026e-05 |
| 2,552 | Updatable Learned Index with Precise Positions | 2021 | VLDB | 8.5530411e-05 |
| 5,157 | Hist-Tree: Those Who Ignore It Are Doomed to Learn | 2021 | CIDR | 5.6589595e-05 |
| 8,811 | Tuning Hierarchical Learned Indexes on Disk and Beyond | 2022 | SIGMOD | 4.4441574e-05 |
| 102 | The Case for Learned Index Structures | 2018 | SIGMOD | 0.00049545203 |
| 7,390 | Making In-Memory Learned Indexes Efficient on Disk | 2024 | SIGMOD | 4.7431654e-05 |
| 1,460 | Benchmarking Learned Indexes | 2021 | VLDB | 0.00011887068 |
| 5,074 | Learned Index: A Comprehensive Experimental Evaluation | 2023 | VLDB | 5.7175726e-05 |
| 6,445 | Updatable Learned Indexes Meet Disk-Resident DBMS - From Evaluations to Design Choices | 2023 | SIGMOD | 5.0589805e-05 |
| 857 | The PGM-index: a fully-dynamic compressed learned index with provable worst-case bounds | 2020 | VLDB | 0.00015882892 |