LAMB: A Training-Free Method to Enhance the Long-Context Understanding of SSMs via Attention-Guided Token Filtering
Zhifan Ye, Zheng Wang, Kejing Xia, Jihoon Hong, Leshu Li, Lexington Whalen, Cheng Wan, Yonggan Fu, Yingyan Celine Lin, Souvik Kundu. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers). 2025.
Authors
- Cheng Wan (ORCID: https://orcid.org/0000-0002-2295-3481)
- Yonggan Fu (ORCID: https://orcid.org/0000-0002-7483-2921)
- L Li
- Souvik Kundu (ORCID: https://orcid.org/0000-0002-1189-3073)
- Zheng Wang (ORCID: https://orcid.org/0000-0001-8305-0676)
- Zhifan Ye (ORCID: https://orcid.org/0000-0001-7705-8719)
- Yingyan Celine Lin
- Jihoon Hong (ORCID: https://orcid.org/0000-0002-2407-8107)
- L. R. Whalen (ORCID: https://orcid.org/0000-0002-9093-0839)
- Kejing Xia
Publication Details
- Published
- 2025-01-01
- DOI
- https://doi.org/10.18653/v1/2025.acl-short.96
- Citations
- 1
- Primary Topic
- Data Stream Mining Techniques
- Type
- article
- Field-Weighted Citation Impact
- 2.48
Funders
- National Science Foundation
- Semiconductor Research Corporation
- Defense Advanced Research Projects Agency