A Unified Adaptive Feature Composition Framework for Multi-Task Generalization in Wireless Foundation Models

2026-06-08

AI Summary

This work addresses the challenges faced by wireless foundation models in multi-task adaptation, where fine-tuning incurs high computational costs and frozen features often lack sufficient representational capacity. To overcome these limitations, the authors propose a unified adaptive feature composition framework that leverages an interpretable routing adapter to extract multi-level features from different depths of a Transformer backbone. A lightweight task-driven network dynamically generates layer-wise aggregation weights, enabling efficient representation fusion without modifying the pre-trained model. Introducing fewer than 50,000 additional parameters, the method significantly outperforms existing adaptation strategies across four representative wireless tasks and reveals interpretable patterns in how different tasks preferentially utilize specific network layers.

Abstract

Though wireless foundation models (WFMs) have shown strong potential in learning universal channel representations, their adaptation to various downstream tasks remains constrained by existing paradigms. Fine-tuning strategies introduces substantial computational and storage overhead, while frozen feature extraction leads to sub-optimal performance across diverse downstream tasks. To address this issue, we propose a unified adaptive feature composition framework for multitask generalization in WFMs, where the key component is the Routing Adapter for Feature Composition (RAFC). Instead of extracting only the final-layer output, this router treats the hidden states from different Transformer depths as a reusable pool of multi-level hidden features, and employs a lightweight task-driven feature composition network to generate layer-wise aggregation weights, then adaptively combine hierarchical representations through weighted summation. This design enables each downstream task to access suitable mixture of low-, mid-, and high-level wireless features without modifying the pretrained backbone. Extensive experiments on four representative wireless tasks demonstrate that RAFC consistently outperforms conventional adaptation baselines while introducing fewer than 50K additional parameters. Moreover, the learned routing weights provide interpretable evidence of task-specific layer preferences, making the proposed framework a low-complexity, scalable, and explainable interface for adapting WFMs to diverse downstream scenarios.

Problem

• wireless foundation models
• multi-task generalization
• feature composition
• downstream task adaptation
• parameter efficiency

Innovation

• adaptive feature composition
• wireless foundation models
• multi-task generalization
• routing adapter
• hierarchical representation