Retrieval-augmented Memory For Online Learning

Abstract

Retrieval-augmented models couple parametric predictors with non-parametric memories, but their use in streaming supervised learning with concept drift is not well understood. We study online classification in non-stationary environments and propose Retrieval-Augmented Memory for Online Learning (RAM-OL), a simple extension of stochastic gradient descent that maintains a small buffer of past examples. At each time step, RAM-OL retrieves a few nearest neighbours of the current input in the hidden representation space and updates the model jointly on the current example and the retrieved neighbours. We compare a naive replay variant with a gated replay variant that constrains neighbours using a time window, similarity thresholds, and gradient reweighting, in order to balance fast reuse of relevant past data against robustness to outdated regimes. From a theoretical perspective, we interpret RAM-OL under a bounded drift model and discuss how retrieval can reduce adaptation cost and improv

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keydu2025retrieval

Related papers

• CREAM: Continual Retrieval On Dynamic Streaming Corpora With Adaptive Soft Memory (2026)0.00
• A Dynamic Retrieval-augmented Generation System With Selective Memory And Remembrance (2026)0.00
• LUMA-RAG: Lifelong Multimodal Agents With Provably Stable Streaming Alignment (2025)0.00
• MURR: Model Updating With Regularized Replay For Searching A Document Stream (2025)0.00
• RAG Without Forgetting: Continual Query-infused Key Memory (2026)0.00
• L^2R: Lifelong Learning For First-stage Retrieval With Backward-compatible Representations (2023)5.24
• RADAR: Recall Augmentation Through Deferred Asynchronous Retrieval (2025)2.26
• Raserec: Retrieval-augmented Sequential Recommendation (2024)2.92