Shallow Probes Are Sufficient for Near-Oracle Personalization Depth Allocation in Federated Learning

arXiv:2610.25193 · cs.LG · Submitted 2026-10-18 · Read on arXiv

cs.LG

Submitted: 2026-10-18

Updated: 2026-10-18

Comments: 6 pages, 9 figures. Accepted and presented at SINCONF 2026 (18th International Conference on Security of Information and Networks), Beykent University, Istanbul, Turkiye, 14-16 October 2026

Journal ref: Proc. 18th Int. Conf. Security of Information and Networks (SINCONF), Istanbul, Turkiye, Oct. 2026

License: http://creativecommons.org/licenses/by-nc-sa/4.0/

The gist: Rather than proposing a new federated optimizer, we study whether a small client-specific measurement budget can support cost-aware personalization-depth decisions under a utility that trades

Terminology

Abstract

Rather than proposing a new federated optimizer, we study whether a small client-specific measurement budget can support cost-aware personalization-depth decisions under a utility that trades Macro-F1 against deployment cost. Exhaustive evaluation of six depths defines an offline oracle that is not deployable. We audit the Probe Sufficiency Hypothesis: three shallow probes (P0--P2) may contain most decision-relevant information for approaching that oracle. FedPBA is a secondary learned instantiation for deep-action extrapolation. On CIFAR-10/100 and patient-disjoint PTB-XL (45 sessions; 900 clients; 2700 evaluations; GroupKFold), training-free greedy probing achieves regret 0.0071 and 98.3 percent oracle recovery, beating metadata-only policies and FedPBA (0.0110 / 97.3 percent). Aggregate shallow-best versus oracle correlation is 0.9818 but falls to 0.578 on deep-oracle rows; PTB-XL deep-oracle cases remain a high-regret boundary without a validated escalation gate.

Related papers