CAFE: Towards Compact, Adaptive, and Fast Embedding for Large-scale Recommendation Models
Summary: CAFE enables compact, adaptive embedding for large-scale DLRMs; HotSketch identifies hot features and assigns them dedicated embeddings, while non-hot features share via multi-level hash. Theoretical accuracy/convergence analysis; 3.92% and 3.68% AUC gains on Criteo Kaggle and CriteoTB at 10k× compression. (summarized by gpt-5-nano on Feb 09 2026)
Incoming Non-self Citations Over Time
Authors
- 1. Hailin Zhang (Peking University)
- 2. Zirui Liu (Peking University)
- 3. Boxuan Chen (Peking University)
- 4. Yikai Zhao (Peking University)
- 5. Tong Zhao (Peking University)
- 6. Tong Yang (Peking University)
- 7. Bin Cui (Peking University)
BibTeX Citation
@inproceedings{zhang_sigmod24,
title = {{CAFE: Towards Compact, Adaptive, and Fast Embedding for Large-scale Recommendation Models}},
author = {Zhang, Hailin and Liu, Zirui and Chen, Boxuan and Zhao, Yikai and Zhao, Tong and Yang, Tong and Cui, Bin},
series = {{SIGMOD} '24},
booktitle = {Proceedings of the {ACM} {SIGMOD} International Conference on Management of Data},
publisher = {Association for Computing Machinery},
doi = {10.1145/3639306},
url = {https://dl.acm.org/doi/10.1145/3639306},
year = {2024}
}Incoming Citations (Sorted by Pagerank)
Showing 2 of 2 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 9,880 | MEMO: Fine-grained Tensor Management For Ultra-long Context LLM Training | 2025 | SIGMOD | 5.2040783e-05 |
| 10,521 | Efficient GNN Training on Giant Graphs with Collective Batching and Scheduling | 2026 | VLDB | 5.093636e-05 |
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 18 of 18 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
Previous
Page 1 / 1
Next