Export Citations
Save this search
Please login to be able to save your searches and receive alerts for new content matching your search criteria.
- research-articleMay 2024
Rethinking the role of token retrieval in multi-vector retrieval
NIPS '23: Proceedings of the 37th International Conference on Neural Information Processing SystemsDecember 2023, Article No.: 677, Pages 15384–15405Multi-vector retrieval models such as ColBERT [Khattab and Zaharia, 2020] allow token-level interactions between queries and documents, and hence achieve state of the art on many information retrieval benchmarks. However, their nonlinear scoring function ...
- research-articleMay 2024
Conditional adapters: parameter-efficient transfer learning with fast inference
- Tao Lei,
- Junwen Bai,
- Siddhartha Brahma,
- Joshua Ainslie,
- Kenton Lee,
- Yanqi Zhou,
- Nan Du,
- Vincent Y. Zhao,
- Yuexin Wu,
- Bo Li,
- Yu Zhang,
- Ming-Wei Chang
NIPS '23: Proceedings of the 37th International Conference on Neural Information Processing SystemsDecember 2023, Article No.: 357, Pages 8152–8172We propose Conditional Adapter (CODA), a parameter-efficient transfer learning method that also improves inference efficiency. CODA generalizes beyond standard adapter approaches to enable a new way of balancing speed and accuracy using conditional ...
- research-articleApril 2024
Mixture-of-experts with expert choice routing
- Yanqi Zhou,
- Tao Lei,
- Hanxiao Liu,
- Nan Du,
- Yanping Huang,
- Vincent Y. Zhao,
- Andrew Dai,
- Zhifeng Chen,
- Quoc Le,
- James Laudon
NIPS '22: Proceedings of the 36th International Conference on Neural Information Processing SystemsNovember 2022, Article No.: 515, Pages 7103–7114Sparsely-activated Mixture-of-experts (MoE) models allow the number of parameters to greatly increase while keeping the amount of computation for a given token or a given sample unchanged. However, a poor expert routing strategy can cause certain experts ...