On the Scaling of Robustness and Effectiveness in Dense Retrieval
Robustness and Effectiveness are critical aspects of developing dense retrieval models for real-world applications. It is known that there is a trade-off between the two. Recent work has addressed scaling laws of effectiveness in dense retrieval, revealing a power-law relationship between effectiveness and the size of models and data. Does robustness follow scaling laws too? If so, can scaling improve both robustness and effectiveness together, or do they remain locked in a trade-off?
Authors
- Yixing Fan (ORCID: https://orcid.org/0000-0003-4317-2702)
- Ruqing Zhang (ORCID: https://orcid.org/0000-0003-4294-2541)
- Xueqi Cheng (ORCID: https://orcid.org/0000-0002-5201-8195)
- Maarten de Rijke (ORCID: https://orcid.org/0000-0002-1086-0202)
- Jiafeng Guo (ORCID: https://orcid.org/0000-0002-9509-8674)
- Yu-An Liu (ORCID: https://orcid.org/0000-0002-9125-5097)
Institutions
- Computer Network Information Center (CN)
- Amsterdam University of the Arts (NL)
- University of Chinese Academy of Sciences (CN)
- University of Amsterdam (NL)
Publication Details
- Published
- 2025-07-13
- DOI
- https://doi.org/10.1145/3726302.3730049
- Primary Topic
- Domain Adaptation and Few-Shot Learning
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- European Commission
- National Natural Science Foundation of China
- Nederlandse Organisatie voor Wetenschappelijk Onderzoek
- Youth Innovation Promotion Association of the Chinese Academy of Sciences
- HORIZON EUROPE Framework Programme
- National Key Research and Development Program of China
- Youth Innovation Promotion Association