Mitigate Negative Transfer with Similarity Heuristic Lifelong Prompt Tuning

Chenyuan Wu, Gangwei Jiang, Defu Lian

Abstract
Lifelong prompt tuning has significantly advanced parameter-efficient lifelong learning with its efficiency and minimal storage demands on various tasks.Our empirical studies, however, highlights certain transferability constraints in the current methodologies: a universal algorithm that guarantees consistent positive transfer across all tasks is currently unattainable, especially when dealing dissimilar tasks that may engender negative transfer.Identifying the misalignment between algorithm selection and task specificity as the primary cause of negative transfer, we present the Similarity Heuristic Lifelong Prompt Tuning (SHLPT) framework. This innovative strategy partitions tasks into two distinct subsets by harnessing a learnable similarity metric, thereby facilitating fruitful transfer from tasks regardless of their similarity or dissimilarity. Additionally, SHLPT incorporates a parameter pool to combat catastrophic forgetting effectively. Our experiments shows that SHLPT outperforms state-of-the-art techniques in lifelong learning benchmarks and demonstrates robustness against negative transfer in diverse task sequences.
Anthology ID:
2024.findings-acl.650
Volume:
Findings of the Association for Computational Linguistics: ACL 2024
Month:
August
Year:
2024
Address:
Bangkok, Thailand
Editors:
Lun-Wei Ku, Andre Martins, Vivek Srikumar
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
10944–10959
Language:
URL:
https://aclanthology.org/2024.findings-acl.650
DOI:
10.18653/v1/2024.findings-acl.650
Bibkey:
Cite (ACL):
Chenyuan Wu, Gangwei Jiang, and Defu Lian. 2024. Mitigate Negative Transfer with Similarity Heuristic Lifelong Prompt Tuning. In Findings of the Association for Computational Linguistics: ACL 2024, pages 10944–10959, Bangkok, Thailand. Association for Computational Linguistics.
Cite (Informal):
Mitigate Negative Transfer with Similarity Heuristic Lifelong Prompt Tuning (Wu et al., Findings 2024)
Copy Citation:
PDF:
https://preview.aclanthology.org/dois-2013-emnlp/2024.findings-acl.650.pdf
PDF Search