# Fast Wireless Foundation Models with Early-Exits

arXiv: https://arxiv.org/abs/2606.29640 · HTML: https://mlsanity.com/paper/2606.29640/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Early exit and edge inference, one of 12 directions in Efficient LLM inference and serving.
- Builds on 11 of its 17 references that are in the map.
- First posted on arXiv 28 Jun 2026.

## Placed in

- Efficient LLM inference and serving / [Early exit and edge inference](https://mlsanity.com/inference/early-exit-and-edge-inference/)

## Builds on (most cited first, 11 of 11)

- [BranchyNet: Fast Inference via Early Exiting from Deep Neural Networks](https://mlsanity.com/paper/1709.01686/) (2017)
- [Layer by Layer: Uncovering Hidden Representations in Language Models](https://mlsanity.com/paper/2502.02013/) (2025)
- [WirelessGPT: A Generative Pre-trained Multi-task Learning Framework for Wireless Communication](https://mlsanity.com/paper/2502.06877/) (2025)
- [Intermediate Layer Classifiers for OOD generalization](https://mlsanity.com/paper/2504.05461/) (2025)
- [6G WavesFM: A Foundation Model for Sensing, Communication, and Localization](https://mlsanity.com/paper/2504.14100/) (2025)
- [A MIMO Wireless Channel Foundation Model via CIR-CSI Consistency](https://mlsanity.com/paper/2502.11965/) (2025)
- [DeepBeam: Deep Waveform Learning for Coordination-Free Beam Management in mmWave Networks](https://mlsanity.com/paper/2012.14350/) (2020)
- [IQFM A Wireless Foundational Model for I/Q Streams in AI-Native 6G](https://mlsanity.com/paper/2506.06718/) (2025)
- [Adaptive Early Exiting for Collaborative Inference over Noisy Wireless Channels](https://mlsanity.com/paper/2311.18098/) (2023)
- [Self-supervised Radio Representation Learning: Can we Learn Multiple Tasks?](https://mlsanity.com/paper/2509.03077/) (2025)
- [Using Early Exits for Fast Inference in Automatic Modulation Classification](https://mlsanity.com/paper/2308.11100/) (2023)

## Built on by (most cited first, 1 of 1)

- [Scaling of Wireless Foundation Models via Representation Diversity and Multi-Branch Architectures](https://mlsanity.com/paper/2610.04289/) (2026)

## Abstract (arXiv)

While wireless foundation models (FMs) are demonstrating strong potential to enable AI-Native 6G networks, their high computational cost remains a critical barrier to deployment. The large computational cost stems from the rigid, full-depth execution of the FM backbone for every task, a process we show is not only inefficient but can also degrade performance on unseen out-of-distribution (OOD) tasks. In this paper, we propose a novel early-exit FM framework that attaches lightweight, per-task heads, at the most appropriate exit-stage of a frozen wireless FM encoder, enabling variable-depth inference tailored to each task's preferred representation depth. Our results demonstrate that these intermediate-layer features not only speed-up inference significantly (up to 93% fewer FLOPs), but also provide more transferable representations that exceed the full encoder accuracy on unseen tasks. We further demonstrate that a simple fixed-exit strategy per task is more effective than traditional early-exiting policies that route different samples to different exits based on their perceived difficulty levels.