Research map / Continual learning / Online and replay-based
Gradient-Guided Epsilon Constraint Method for Online Continual Learning
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Revisits experience replay for online continual learning through epsilon-constraint optimization and proposes a gradient-guided epsilon-constraint method.
Problem
Online continual learning with limited memory, where rehearsal-based methods such as Experience Replay (ER) are commonly used but their implicit handling of past-task performance is not well understood.
Approach
Reinterprets ER as an implicit soft epsilon-constraint on past-task performance, where the weighting parameter post-hoc defines a slack variable; builds a gradient-guided epsilon-constraint method on this view.
Experimental setup
Baselines: Experience Replay
Builds on: Experience Replay
Closest papers by meaning
- Online Curvature-Aware Replay: Leveraging $\mathbf{2^{nd}}$ Order Information for Online Continual Learning (ICML 2025 · similarity 0.91)
- Mitigating Catastrophic Forgetting in Online Continual Learning by Modeling Previous Task Interrelations via Pareto Optimization (ICML 2024 · similarity 0.89)
- Layerwise Proximal Replay: A Proximal Point Method for Online Continual Learning (ICML 2024 · similarity 0.89)
- Advancing Prompt-Based Methods for Replay-Independent General Continual Learning (ICLR 2025 · similarity 0.89)
- Summarizing Stream Data for Memory-Constrained Online Continual Learning (AAAI 2024 · similarity 0.88)
- Enhancing Online Continual Learning with Plug-and-Play State Space Model and Class-Conditional Mixture of Discretization (CVPR 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.