This work introduces a novel measure, the relative parameter-importance, which measures the relative importance of each parameter with respect to both the current and past tasks, and allows the update of parameters with high past-task importance when they have low relative-importance, thus enabling backward knowledge transfer.
Abstract
Achieving continual learning (CL) with deep neural networks requires balancing stability and plasticity while enabling knowledge transfer. In this work, we focus on offline learning algorithms under the constraints: (I) no access to training data from prior tasks (II) no access to task-id at inference time. We introduce a novel measure, the relative parameter-importance, which measures the relative importance of each parameter with respect to both the current and past tasks. Parameters with high relative importance are interpreted as more important for maintaining past-task stability and thus heavily regularised, whereas parameters with low relative-importance are allowed to be more freely updated. Unlike existing methods, our approach allows the update of parameters with high past-task importance when they have low relative-importance, thus enabling backward knowledge transfer in addition to tackling the stability-plasticity trade-off. We demonstrate improvements against state-of-the-art CL methods on both class-incremental and domain-incremental learning text classification problems and provide insights for extending our method to text generation problems. Code available at: https://github.com/itsmemala/LACL
Experiments show that FiUni can effectively infer latent batch-level task affiliations and achieve competitive performance against advanced task-aware CL methods with fewer trainable parameters.
Dezheng Han, Anlan Zhang, Zhiwu Zhu et al.· 0 citations
Continual learning seeks to enable a model to assimilate knowledge from non-stationary data streams without catastrophic forgetting. Recently, methods based on Parameter-Efficient Tuning (PET) have achieved superior performance without even storing any historical exemplars, which train much fewer specific parameters for each task upon a frozen pre-trained model, and tailored parameters are retrieved to guide predictions during inference. However, reliance solely on pre-trained features for parameter matching exacerbates the inconsistency between the training and inference phases, thereby constraining the overall performance. To address this issue, we propose HRM-PET, which makes full use of the richer downstream knowledge inherently contained in the trained parameters. Specifically, we introduce a hybrid re-matching mechanism, which benefits from the initial predicted distribution to facilitate the parameter selections. The direct re-matching addresses misclassified samples identified with correct task identity in prediction, despite incorrect initial matching. Moreover, the confidence-based re-matching is specifically designed to handle other more challenging mismatched samples that cannot be calibrated by the former. Besides, to acquire task-invariant knowledge for better matching, we integrate a cross-task instance relationship distillation module into the PET-based method. Extensive experiments conducted on four datasets under five pre-trained settings demonstrate that HRM-PET performs favorably against the state-of-the-art methods. The code is available
Weicheng Wang, Guoli Jia, Xialei Liu et al.· Neural Information Processin...· 0 citations
Artificial general intelligence requires neural networks to continually acquire new knowledge while preserving previously learned information. However, sequential learning causes catastrophic forgetting, whereby performance on earlier tasks deteriorates as new tasks are learned. Parameter regularization methods, such as Elastic Weight Consolidation (EWC), attempt to address this challenge by constraining updates to parameters deemed important for previous tasks. This work investigates the influence of fully connected (FC) layer architecture on parameter regularization in the class incremental learning setting using a modified ResNet-18 trained on the CIFAR-10 dataset. Experiments demonstrate that a dynamically expanding FC architecture substantially outperforms a conventional static FC layer. To address the limited memory retention observed with EWC in static FC architectures, a new parameter regularization approach, termed Range Matrix, is proposed. Unlike EWC, which selectively constrains parameters identified by the Fisher Information Matrix, Range Matrix constrains all network parameters within dynamically computed bounds derived from loss gradients. Experimental results show that Range Matrix improves memory retention and reduces catastrophic forgetting relative to EWC when using a static FC architecture. Control experiments further demonstrate that the improved performance of Dynamic FC arises primarily from context partitioning rather than increased model capacity alone. Finally, analysis of the overlap between Fisher-important parameters and Range Matrix-constrained parameters shows that the two methods identify substantially different parameter subsets despite achieving similar performance under Dynamic FC, suggesting that FC architecture is the dominant factor governing continual learning performance. These findings provide both a novel parameter regularization strategy and new insights into the interaction between network architecture and catastrophic forgetting in continual learning.
Henry Huang· Journal of high school scien...· 0 citations
Low-Rank Adaptation (LoRA) enables foundation models to efficiently adapt to downstream tasks with minimal trainable parameters, yet it lacks the capability of continual learning (CL) from non-stationary data streams. While existing CL methods attempt to address this by storing and retrieving task-specific parameter modules, they suffer from non-trivial challenges of task identification and inconsistency in input-output feature mappings, resulting in suboptimal feature representations. Instead of keeping multiple task-specific parameter modules, we propose a novel continual LoRA mechanism that parameterizes all task-specific adaptations with a single adaptation module, naturally obviating the task identification and mapping inconsistency issues. To this end, we reformulate LoRA-based CL as a consistent feature mapping problem that mimics the behavior of the joint-training upper bound, wherein a unified adaptation parameter matrix is learned to simultaneously capture the input-output relationships established by all task-specific LoRAs. By solving this formulation, our approach facilitates approximate unified optimization over sequential tasks via progressively updating cumulative feature statistics, achieving efficient knowledge retention and renewal. Notably, the proposed arithmetic training mechanism can be applied both to the backbone of the foundation model for parameter-efficient fine-tuning and the classifier optimization, thereby enabling continual learning within an integrated architecture that closely approximates the joint learning paradigm. Extensive experiments across conventional and long-sequence class-incremental benchmarks demonstrate that our method achieves new state-of-the-art performance.
Yue Lu, Shizhou Zhang, De Cheng et al.· IEEE Transactions on Pattern...· 0 citations
Deep neural networks excel in various tasks but struggle to generalize across evolving data distributions, leading to significant performance degradation under domain shifts. Domain incremental learning (DIL) addresses this challenge by enabling models to continuously adapt while retaining prior knowledge. Among existing DIL approaches, the parameter-isolation paradigm achieves state-of-the-art performance. However, these methods often adopt a one-size-fits-all approach to adapt to new domains, resulting in either insufficient learning capacity or redundant parameters. In this work, we propose BPG, a unified framework that addresses both challenges through two complementary components: BPG-Adapter, which dynamically determines each domain's adapter hidden dimension based on domain-specific feature separability, and BPG-Inference, a soft domain mixture strategy that integrates multiple domain-specific models at test time, mitigating domain ID misselection. Experimental results on DomainNet, CDDB, and CORe50 demonstrate that BPG consistently outperforms uniform adapter-based approaches and hard domain selection strategies, achieving state-of-the-art average accuracy while reducing forgetting to as low as 0.22% on DomainNet.
Qiang Wang, Songlin Dong, Shaokun Wang et al.· 0 citations
This work formalizes CL as an online optimization problem governed by the interaction between environmental and learning dynamics, and introduces Transfer Efficiency as a quantitative measure of the tension between Instability, the bias inherited from conflicting past experience, and Transient Error, the optimization cost of learning new tasks from scratch.
Giulia Lanzillotta, Mandana Samiei, D. Precup et al.· 0 citations
