연구
DEAF: A Benchmark for Diagnostic Evaluation of Acoustic Faithfulness in Audio Language Models
arXiv:2603.18048v1 Announce Type: new Abstract: Recent Audio Multimodal Large Language Models Audio MLLMs demonstrate impressive performance on speech benchmarks, yet it remains unclear whether these models genuinely process acoustic signals or rely on textbased semantic inference.
This content is a summary of the original article from ArXiv AI. Please visit the original site for the full article.
Read Original Article →