Dynamic Task Vector Grouping for Efficient Multi-Task Prompt Tuning

2025-03-23

AI Summary

This work addresses the challenge of selecting optimal source-task combinations for low-resource target tasks in multi-task prompt tuning, where static similarity estimation and fixed transfer strategies often induce negative transfer. To mitigate this, we propose a dynamic task vector grouping mechanism that jointly leverages target-task similarity and knowledge consistency to perform real-time clustering and adaptive updating of source-task groups during fine-tuning—enabling adaptive soft-prompt transfer and similarity-aware reweighting. Unlike conventional single-task or all-task transfer paradigms, our approach is the first to formulate task composition as a time-varying optimization problem. Extensive experiments across 26 NLP datasets demonstrate that our method effectively suppresses negative transfer and achieves state-of-the-art performance.

Technology Category

• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Natural Language Processing: Prompt Engineering / Prompting
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• User Modeling, Personalization and Recommendation: User modeling for targeted and personalized online advertising
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work

Abstract

Multi-task prompt tuning utilizes multiple high-resource source tasks to improve performance on low-source target tasks. Existing approaches transfer the soft prompt trained by combining all source tasks or a single ``high-similar'' source task one-time-only. However, we find that the optimal transfer performance often comes from a combination of source tasks, which is neither one nor all. Further, we find that the similarity between source and target tasks also changes dynamically during fine-tuning after transfering, making similarity calculation in the initiation stage inadequate. To address these issues, we propose a method called Dynamic Task Vector Grouping (DTVG), whose core ideas contain (1) measuring the task similarity with task vectors instead of soft prompt, (2) grouping the optimal source task combination based on two metrics: {it target similarity} and {it knowledge consistency}; (3) dynamically updating the combination in each iteration step. Extensive experiments on the 26 NLP datasets under different settings demonstrate that DTVG effectively groups similar source tasks while reducing negative transfer, achieving the start-of-art performance.

Problem

• Optimal source task combination for transfer learning
• Dynamic similarity between source and target tasks
• Negative transfer reduction in multi-task prompt tuning

Innovation

• Dynamic task vector grouping for optimal transfer
• Task similarity measured with task vectors
• Dynamic source task combination updates