Bias in the Ear of the Listener: Assessing Sensitivity in Audio Language Models Across Linguistic, Demographic, and Positional Variations (2026 · 1 citation)
Audio-capable multimodal LLMs now answer spoken questions directly. It was unclear whether their answers change depending on the speaker's gender, the language or accent, or the order of the answer options. The authors built BiasInEar, a spoken multiple-choice benchmark derived from Global MMLU Lite in English…
Do Multimodal LLMs Understand Order? Measuring the Fragility of Multimodal Reasoning under Input Order Perturbations (2026)
Large vision-language models (LVLMs) do well on multimodal benchmarks, but it is unclear whether their answers stay stable when the same input is arranged differently.

All works

3 works of theirs in our library, the most cited first; 2 with a page here, the others link their original until we summarise them.

Bias in the Ear of the Listener: Assessing Sensitivity in Audio Language Models Across Linguistic, Demographic, and Positional VariationsFindings of the Association for Computational LinguisticsSheng-Lun Wei, Yu-Ling Liao, Yen-Hua Chang, et al.2026 · 1 citation
Do Multimodal LLMs Understand Order? Measuring the Fragility of Multimodal Reasoning under Input Order PerturbationsLanguage Resources and Evaluation Conference (LREC)Sheng-Lun Wei, Yu-Ling Liao, Hen-Hsen Huang, et al.2026
No One Fits All: From Fixed Prompting to Learned Routing in Multilingual LLMs ↗Findings of the Association for Computational Linguistics: ACL 2026 · no summary yetWei-Chi Wu et al.2026

Co-authors here

3 co-authors with a page, the most papers together first.

Where their works here were published, the most works first, with the editions of those that have a page.

Updated 1 November 2026