MetaStore: Analyzing Deep Learning Meta-Data at Scale
Summary: MetaStore stores compact backprop intermediates—prefix and suffix gradients—that exactly reconstruct full model gradients, drastically reducing gradient size. It runs gradient analytics directly on these compact structures, achieving 4–678× storage and 2–1000× runtime gains on VGG/BERT/ResNet. (summarized by gpt-5-mini on Feb 09 2026)
Incoming Non-self Citations Over Time
No non-self incoming citations found for this paper in this database.
Authors
- 1. Huayi Zhang
- 2. Binwei Yan
- 3. Lei Cao
- 4. Samuel Madden
- 5. Elke Rundensteiner
Incoming Citations (Sorted by Pagerank)
Showing 0 of 0 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 6 of 6 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
| Rank | Cited Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 2,157 | MISTIQUE: A System to Store and Query Model Intermediates for Model Diagnosis | 2018 | SIGMOD | 9.4153917e-05 |
| 2,759 | Complaint-driven Training Data Debugging for Query 2.0 | 2020 | SIGMOD | 8.1646193e-05 |
| 4,553 | A Demonstration of AutoOD: A Self-Tuning Anomaly Detection System | 2022 | VLDB | 6.0852762e-05 |
| 6,001 | FACET: Robust Counterfactual Explanation Analytics | 2023 | SIGMOD | 5.2365238e-05 |
| 6,004 | DeepEverest: Accelerating Declarative Top-K Queries for Deep Neural Network Interpretation | 2022 | VLDB | 5.2365238e-05 |
| 8,713 | LANCET: Labeling Complex Data at Scale | 2021 | VLDB | 4.4577046e-05 |
Previous
Page 1 / 1
Next