Continual Learning

VQ-Prompt uses vector quantization to optimize discrete prompts for continual learning, achieving state-of-the-art performance by effectively abstracting task knowledge and optimizing prompt selection…

AdaGauss tackles task-recency bias in exemplar-free class incremental learning by adapting class covariances and introducing an anti-collapse loss, achieving state-of-the-art results.

SAFE, a novel parameter-efficient tuning framework, boosts pre-trained model performance in continual learning by balancing model stability and plasticity through slow and fast learning stages, signif…

Forget-free graph class-incremental learning achieved via a novel task profiling and prompting approach, significantly outperforming state-of-the-art methods.

Model Sensitivity Aware Continual Learning (MACL) tackles the CL challenge by optimizing model performance based on parameter distribution, achieving superior old knowledge retention and new task perf…

Boost continual learning with C-Flat: a novel, one-line-code optimizer creating flatter loss landscapes for enhanced stability and generalization across various continual learning scenarios.

Bridging the accuracy gap in online continual learning caused by label delays, a new framework with Importance Weighted Memory Sampling prioritizes relevant memory samples, significantly outperforming…

GACL: a novel exemplar-free technique for generalized analytic continual learning, achieves superior performance by analytically solving the weight-invariant property for handling real-world data.

NsCE framework tackles key OCL challenges: model ignorance (learning effective features in limited time) and myopia (overly simplified features). NsCE integrates non-sparse maximum separation regulari…

Boost continual learning efficiency with CLFD: a novel frequency domain approach that improves accuracy by up to 6.83% and slashes training time by 2.6x on edge devices!