Scaling Up Temporal Domain Generalization via Temporal Experts Averaging
This addresses the challenge of generalizing across temporal distribution shifts in machine learning, offering a scalable solution for applications like NLP, though it is incremental as it builds on prior weight prediction methods.
The paper tackles the problem of temporal domain generalization (TDG) by proposing Temporal Experts Averaging (TEA), a scalable framework that updates entire models via weight averaging to handle temporal shifts like lexical change, achieving up to 69% performance improvement and 60x efficiency gains over prior methods.
Temporal Domain Generalization (TDG) aims to generalize across temporal distribution shifts, e.g., lexical change over time. Prior work often addresses this by predicting future model weights. However, full model prediction is prohibitively expensive for even reasonably sized models. Thus, recent methods only predict the classifier layer, limiting generalization by failing to adjust other model components. To address this, we propose Temporal Experts Averaging (TEA), a novel and scalable TDG framework that updates the entire model using weight averaging to maximize generalization potential while minimizing computational costs. Our theoretical analysis guides us to two steps that enhance generalization to future domains. First, we create expert models with functional diversity yet parameter similarity by fine-tuning a domain-agnostic base model on individual temporal domains while constraining weight changes. Second, we optimize the bias-variance tradeoff through adaptive averaging coefficients derived from modeling temporal weight trajectories in a principal component subspace. Expert's contributions are based on their projected proximity to future domains. Extensive experiments across 7 TDG benchmarks, 5 models, and 2 TDG settings shows TEA outperforms prior TDG methods by up to 69% while being up to 60x more efficient.