Task2vec Readiness: Diagnostics for Federated Learning from Pre-Training Embeddings

arXiv:2604.10849v1 Announce Type: cross Abstract: Federated learning (FL) performance is highly sensitive to heterogeneity across clients, yet practitioners lack reliable methods to anticipate how a federation will behave before training. We propose readiness indices, derived from Task2Vec embeddings, that quantifies the alignment of a federation prior to training and correlates with its eventual performance. Our approach computes unsupervised metrics -- such as cohesion, dispersion, and density -- directly from client embeddings. We evaluate these indices across diverse datasets (CIFAR-10, FEMNIST, PathMNIST, BloodMNIST) and client counts (10--20), under Dirichlet heterogeneity levels spanning alpha in {0.05,ots,5.0} and FedAVG aggregation strategy. Correlation analyses show consistent and significant Pearson and Spearman coefficients between some of the Task2Vec-based readiness and final performance, with values often exceeding 0.9 across datasetimesclient configurations, validating this approach as a robust proxy for FL outcomes. These findings establish Task2Vec-based readiness as a principled, pre-training diagnostic for FL that may offer both predictive insight and actionable guidance for client selection in heterogeneous federations.

Related

• From Selection to Scheduling: Federated Geometry-Aware Correction Makes Exemplar Replay Work Better under Continual Dynamic Heterogeneity
• Face Density as a Proxy for Data Complexity: Quantifying the Hardness of Instance Count
• Fairness is Not Flat: Geometric Phase Transitions Against Shortcut Learning
• Learning When Not to Learn: Risk-Sensitive Abstention in Bandits with Unbounded Rewards

Source: arXiv cs.AI | 2026-04-14