Forget to Generalize: Iterative Adaptation for Generalization in Federated Learning DOI: XXXXXXX.XXXXXXX Conference: The Web Conference 2026; June 03–05, 2018; Woodstock, NY ISBN: 978-1-4503-XXXX-X/2018/06 CCS: Computing methodologies Distributed computing methodologies

Abdulrahman Alotaibi, Irene Tenison, Miriam Kim, Isaac Lee, Lalana Kagal · 2026

Read Forget to Generalize: Iterative Adaptation for Generalization in Federated Learning DOI: XXXXXXX.XXXXXXX Conference: The Web Conference 2026; June 03–05, 2018; Woodstock, NY ISBN: 978-1-4503-XXXX-X/2018/06 CCS: Computing methodologies Distributed computing methodologies by Abdulrahman Alotaibi, Irene Tenison, Miriam Kim, Isaac Lee, Lalana Kagal (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-04536.txt.

Description

Abdulrahman Alotaibi and 4 co-authors propose a training trick for federated learning in a 2026 paper. Federated learning trains one model across many devices with no central gathering of private data. It struggles when each device's data differ greatly from the others.

Their method, Iterative Federated Adaptation, splits training into generations. At each generation's end, some model parameters are reset to fresh values. The reset parameters are picked at random or from later layers. Forgetting in this controlled way helps the model escape poor fits and keep broadly useful traits. Tests on three image datasets raised accuracy, most when data differed strongly between clients.

Catalog record

CREATORAbdulrahman Alotaibi, Irene Tenison, Miriam Kim, Isaac Lee, Lalana Kagal
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE22 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2602-04536

Browse the full library · About the Tower of Babel