[2603.02464] GLoRIA: Gated Low-Rank Interpretable Adaptation for Dialectal ASR

Computer Science > Computation and Language arXiv:2603.02464 (cs) [Submitted on 2 Mar 2026] Title:GLoRIA: Gated Low-Rank Interpretable Adaptation for Dialectal ASR Authors:Pouya Mehralian, Melissa Farasyn, Anne Breitbarth, Anne-Sophie Ghyselen, Hugo Van hamme View a PDF of the paper titled GLoRIA: Gated Low-Rank Interpretable Adaptation for Dialectal ASR, by Pouya Mehralian and 4 other authors View PDF HTML (experimental) Abstract:Automatic Speech Recognition (ASR) in dialect-heavy settings remains challenging due to strong regional variation and limited labeled data. We propose GLoRIA, a parameter-efficient adaptation framework that leverages metadata (e.g., coordinates) to modulate low-rank updates in a pre-trained encoder. GLoRIA injects low-rank matrices into each feed-forward layer, with a gating MLP determining the non-negative contribution of each LoRA rank-1 component based on location metadata. On the GCND corpus, GLoRIA outperforms geo-conditioned full fine-tuning, LoRA, and both dialect-specific and unified full fine-tuning, achieving state-of-the-art word error rates while updating under 10% of parameters. GLoRIA also generalizes well to unseen dialects, including in extrapolation scenarios, and enables interpretable adaptation patterns that can be visualized geospatially. These results show metadata-gated low-rank adaptation is an effective, interpretable, and efficient solution for dialectal ASR. Comments: Subjects: Computation and Language (cs.CL); Artificia...