# Federated Learning Under Temporal Drift -- Mitigating Catastrophic Forgetting via Experience Replay

arXiv: https://arxiv.org/abs/2601.13456 · HTML: https://mlsanity.com/paper/2601.13456/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 4 of its 5 references that are in the map.
- First posted on arXiv 19 Jan 2026.

## Builds on (most cited first, 4 of 4)

- [Communication-Efficient Learning of Deep Networks from Decentralized Data](https://mlsanity.com/paper/1602.05629/) (2016)
- [Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms](https://mlsanity.com/paper/1708.07747/) (2017)
- [Deep Leakage from Gradients](https://mlsanity.com/paper/1906.08935/) (2019)
- [Concept drift detection and adaptation for federated and continual learning](https://mlsanity.com/paper/2105.13309/) (2021)

## Abstract (arXiv)

Federated Learning struggles under temporal concept drift where client data distributions shift over time. We demonstrate that standard FedAvg suffers catastrophic forgetting under seasonal drift on Fashion-MNIST, with accuracy dropping from 74% to 28%. We propose client-side experience replay, where each client maintains a small buffer of past samples mixed with current data during local training. This simple approach requires no changes to server aggregation. Experiments show that a 50-sample-per-class buffer restores performance to 78-82%, effectively preventing forgetting. Our ablation study reveals a clear memory-accuracy trade-off as buffer size increases.