Metraipolis › archive › dc_fe298eb7962b
Pruning for Efficiency, Paying in Fairness: Demographic Disparities in Pruned Speech-LLMs — arXiv:2609.38106
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.38106
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T16:57:45.064Z
- SHA-256 of the archived text
b0e4879fd464ac85463ba356663ee4f80ca480f09bec7b275b0de3cb794b95f7- Stable URL of this page
https://metraipolis.com/documento/dc_fe298eb7962b
How to cite
Pruning for Efficiency, Paying in Fairness: Demographic Disparities in Pruned Speech-LLMs — arXiv:2609.38106 — arXiv (en), https://arxiv.org/abs/2609.38106 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_fe298eb7962b (Metraipolis archive dc_fe298eb7962b, sha256 b0e4879fd464ac85463ba356663ee4f80ca480f09bec7b275b0de3cb794b95f7)
BibTeX
@misc{dcfe298eb7962b,
title = {Pruning for Efficiency, Paying in Fairness: Demographic Disparities in Pruned Speech-LLMs — arXiv:2609.38106},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.38106},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_fe298eb7962b; sha256 b0e4879fd464ac85463ba356663ee4f80ca480f09bec7b275b0de3cb794b95f7},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.38106. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Speech-LLMs are expensive to run, making compression important for real-world deployment. However, compressed models are usually selected using aggregate word error rate (WER), which can hide how pruning affects different demographic groups. In this work, we systematically study the effect of audio encoder pruning on SLAM-ASR for different demographic groups. Using the Fair-Speech and Common Voice datasets, we found that the pruning does not affect all demographic groups equally; the gap between best- and worst-performing groups increases in fold. These disparities appear across all three encoder scales, but only the largest model initially hides them behind aggregate WER. LoRA adaptation improves WER for every group, but benefits groups already performing well more strongly and widens for certain groups. On Common Voice English, Danish, and Dutch, accent gaps persist but do not clearly widen, showing that the fairness effects of pruning vary across datasets and must be measured directly. Our findings suggest that for pruned models, deployment decisions should include per-group WER, with the worst-performing group's error rate as an explicit criterion. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-30 in cs.CL). Categorie arXiv: Sound (cs.SD) ; Computation and Language (cs.CL). Pagina del lavoro: https://arxiv.org/abs/2609.38106 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-07T16:27:03.472Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-30?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).