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

Source: [Federated Learning Under Temporal Drift -- Mitigating Catastrophic Forgetting via Experience Replay](<https://www.sota2.com/research/paper/69705de7e25ad2fa9d1ed776>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-01-22T05:00:38.183Z

Authors: Sahasra Kokkula, Daniel David, Aaditya Baruah

Published: 2026-01-19T23:17:10.000Z

## Abstract

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.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Image Classification on Fashion-MNIST (seasonal setup)](<https://www.sota2.com/research/sota/image-classification-on-fashion-mnist-seasonal-setup.md>): Classification; Fashion-MNIST seasonal setup; Final Accuracy: 78.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2601.13456>)

[PDF](<https://arxiv.org/pdf/2601.13456.pdf>)
