# Activation Steering for Accent Adaptation in Large Audio Language Models

arXiv: https://arxiv.org/abs/2603.05813 · HTML: https://mlsanity.com/paper/2603.05813/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 6 Mar 2026.

## Built on by (most cited first, 4 of 4)

- [SALSA: Speech Aware LLM Adaptation via Learned Steering Activation Vectors](https://mlsanity.com/paper/2606.00460/) (2026)
- [Encoder-Side Neuron Identification and Amplification for Acoustic Perception in Large Audio-Language Models](https://mlsanity.com/paper/2607.11801/) (2026)
- [Controlling Speaking Rate in Autoregressive TTS via Activation Steering](https://mlsanity.com/paper/2609.33810/) (2026)
- [Steering Speech-Language Models: Training-Free Task Specialization via Contrastive Activation Addition](https://mlsanity.com/paper/2610.04683/) (2026)

## Abstract (arXiv)

Accent variability remains a major source of errors in automatic speech recognition, yet most adaptation methods rely on parameter fine-tuning without understanding where accent information is encoded. We treat accent variation as an interpretable subspace in hidden representations and investigate whether it can be identified and controlled directly in activation space. We extract layer-wise encoder activations and estimate mean-shift directions capturing accent-induced representation shifts. By injecting these directions into individual layers and measuring how they align accented and standard embeddings, we derive a layer-wise accent sensitivity profile, revealing that accent information concentrates in a narrow band of middle encoder layers. Leveraging this structure, we further introduce parameter-free accent steering that modifies representations during inference without updating model weights. Experiments across eight accents show consistent word error rate reductions.