Large pre-trained speech models such as Whisper offer strong generalization but pose significant challenges for resource-efficient adaptation. Low-Rank Adaptation (LoRA) has become a popular parameter-efficient fine-tuning method, yet its underlying mechanisms in speech tasks remain poorly understood. In this work, we conduct the first systematic mechanistic interpretability study of LoRA within the Whisper encoder for speech emotion recognition (SER). Using a suite of analytical tools, including layer contribution probing, logit-lens inspection, and representational similarity via singular value decomposition (SVD) and centered kernel alignment (CKA), we reveal two key mechanisms: a delayed specialization process that preserves general features in early layers before consolidating task-specific information, and a forward alignment, backward differentiation dynamic between LoRA's matrices. Our findings clarify how LoRA reshapes encoder hierarchies, providing both empirical insights and a deeper mechanistic understanding for designing efficient and interpretable adaptation strategies in large speech models. Our code is available at https://github.com/harryporry77/Behind-the-Scenes.

翻译：诸如Whisper等大规模预训练语音模型虽具备强大的泛化能力，但其资源高效适配仍面临重大挑战。低秩适配（LoRA）已成为一种流行的参数高效微调方法，然而其在语音任务中的底层机制仍鲜为人知。本研究首次针对Whisper编码器在语音情感识别任务中的LoRA机制进行了系统性可解释性分析。通过采用层贡献探测、对数透镜检测以及基于奇异值分解（SVD）和中心核对齐（CKA）的表征相似性分析等系列工具，我们揭示了两个关键机制：一是延迟特化过程——该过程在早期层保留通用特征，随后才整合任务特定信息；二是LoRA矩阵间存在前向对齐、后向分化的动态特性。我们的发现阐明了LoRA如何重塑编码器层级结构，为设计大规模语音模型的高效可解释适配策略提供了实证依据与深层机理理解。代码已开源：https://github.com/harryporry77/Behind-the-Scenes。