#natural language process...
Oct 2025
Hearing the Order: Investigating Selection Bias in Large Audio-Language Models
This work demonstrates that no model is immune to this bias through extensive experiments on six LALMs across three widely used benchmarks and their spoken counterparts, and study permutation-based strategies and show that they can mitigate bias in most cases.
Yu-Xiang Lin, Chen-An Li, Sheng-Lun Wei et al.
· arXiv.org · 6 citations