GAP-Prompt: Gated Adaptive Prompting for Efficient Continual Learning

arXiv:2608.23782v1 Announce Type: new Abstract: Continual learning faces the persistent challenge of catastrophic forgetting, where sequential task updates degrade previously acquired knowledge. While prompt-based methods integrated with pre-trained models offer a compelling solution by freezing the backbone, they often rely on static, task-level prompting strategies that overlook fine-grained intra-task diversity. In this paper, we propose Gated Adaptive Prompting (GAP-Prompt), a novel method that introduces instance-level adaptability to the prompting process. GAP-Prompt consists of three synergistic modules: (1) instance-conditioned gating, which dynamically determines optimal prompt injection layers for each individual image; (2) dynamic knowledge fusion, which performs instance-aware aggregation of current and historical prompts, enabling knowledge integration across tasks; and (3) shared prompt distillation, which anchors foundational knowledge in early shared layers to mitigate forgetting. Extensive evaluations on CIFAR-100, ImageNet-R, and CUB-200 benchmarks demonstrate that GAP-Prompt consistently achieves state-of-the-art performance. Notably, on the fine-grained CUB-200 dataset, GAP-Prompt reaches 87.29% accuracy, approaching the joint training upper bound (88.00%) and outperforming existing methods by a significant margin.

Related

• PMF-CL: Pareto-Minimal-Forgetting Continual Learner for Conflicting Tasks
• SAE-FD: Sparse Autoencoder Feature Distillation for Continual Learning of Large Language Models
• Fine-Tuning Without Forgetting via Loss-Adaptive Learning Rates
• Reliability-Gated Source Anchoring for Continual Test-Time Adaptation

Source: arXiv cs.LG | 2026-08-26