Toward Enhancing Representation Learning in Federated Multi-Task Settings

Mehdi Setayesh, Mahdi Beitollahi, Yasser H. Khalil, and Hongliang Li Huawei Noah’s Ark Lab, Montreal, Canada [email protected] [email protected] [email protected] [email protected] · 2026

Read Toward Enhancing Representation Learning in Federated Multi-Task Settings by Mehdi Setayesh, Mahdi Beitollahi, Yasser H. Khalil, and Hongliang Li Huawei Noah’s Ark Lab, Montreal, Canada [email protected] [email protected] [email protected] [email protected] (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-2602-01626.txt.

Description

Mehdi Setayesh and colleagues improve federated multi task learning, where users with different tasks train models together while keeping data private. Most methods assume models match across users, which fails in realistic mixed settings. Instead of sharing model parameters, the authors learn a representation space across tasks. Their Muscle loss aligns representations from all models at once, unlike pairwise contrastive methods, and its minimization matches maximizing mutual information across model representations. The resulting FedMuscle algorithm handles model and task diversity efficiently, beating baselines on varied image and language tasks.

Catalog record

CREATORMehdi Setayesh, Mahdi Beitollahi, Yasser H. Khalil, and Hongliang Li Huawei Noah’s Ark Lab, Montreal, Canada [email protected] [email protected] [email protected] [email protected]
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE104736
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2602.01626

Browse the full library · About the Tower of Babel