Teaching Prompts to Coordinate: Hierarchical Layer-Grouped Prompt Tuning for Continual Learning

TL;DR

This paper introduces a hierarchical layer-grouped prompt tuning approach for continual learning, enhancing stability and reducing catastrophic forgetting by sharing prompts within groups and using a root prompt for sub-prompt generation.

Contribution

It proposes a novel hierarchical prompt tuning method that improves stability and performance in continual learning by sharing prompts within groups and conditioning sub-prompts on a root prompt.

Findings

Achieves favorable performance on four benchmarks.

Reduces catastrophic forgetting compared to existing methods.

Enhances model stability through hierarchical prompt sharing.

Abstract

Prompt-based continual learning methods fine-tune only a small set of additional learnable parameters while keeping the pre-trained model's parameters frozen. It enables efficient adaptation to new tasks while mitigating the risk of catastrophic forgetting. These methods typically attach one independent task-specific prompt to each layer of pre-trained models to locally modulate its features, ensuring that the layer's representation aligns with the requirements of the new task. However, although introducing learnable prompts independently at each layer provides high flexibility for adapting to new tasks, this overly flexible tuning could make certain layers susceptible to unnecessary updates. As all prompts till the current task are added together as a final prompt for all seen tasks, the model may easily overwrite feature representations essential to previous tasks, which increases the risk of catastrophic forgetting. To address this issue, we propose a novel hierarchical layer-grouped prompt tuning method for continual learning. It improves model stability in two ways: (i) Layers in the same group share roughly the same prompts, which are adjusted by position encoding. This helps preserve the intrinsic feature relationships and propagation pathways of the pre-trained model within each group. (ii) It utilizes a single task-specific root prompt to learn to generate sub-prompts for each layer group. In this way, all sub-prompts are conditioned on the same root prompt, enhancing their synergy and reducing independence. Extensive experiments across four benchmarks demonstrate that our method achieves favorable performance compared with several state-of-the-art methods.