新研究表明，大型音频语言模型（LALMs）可能不具备真正的听觉感知能力，尽管它们在基准测试中得分很高。研究显示，这些模型仅凭文本和通用知识就能回答问题，在没有音频输入的情况下仍能保持相当一部分性能。此外，当需要音频时，模型通常只需要片段而非完整音频剪辑，这挑战了当前评估方法在衡量稳健音频理解方面的可靠性。 AI
影响 挑战了当前音频语言模型的评估指标，表明需要更稳健的基准设计来准确衡量听觉理解能力。
排序理由 该集群包含两篇在arXiv上发表的关于大型音频语言模型评估的学术论文。
AI 生成摘要 · Google Gemini · 来自 4 个来源。 我们如何撰写摘要 →
arXiv:2604.24401v1 Announce Type: cross Abstract: Large Audio-Language Models show consistent performance gains across speech and audio benchmarks, yet high scores may not reflect true auditory perception. If a model can answer questions without processing the acoustic signal, th…
arXiv:2510.00626v3 Announce Type: replace-cross Abstract: Large audio-language models (LALMs) unify speech and text processing, but their robustness in noisy real-world settings remains underexplored. We investigate how irrelevant audio, such as silence, synthetic noise, and envi…
Large Audio-Language Models show consistent performance gains across speech and audio benchmarks, yet high scores may not reflect true auditory perception. If a model can answer questions without processing the acoustic signal, the benchmark fails as a measure of auditory underst…
Large Audio-Language Models show consistent performance gains across speech and audio benchmarks, yet high scores may not reflect true auditory perception. If a model can answer questions without processing the acoustic signal, the benchmark fails as a measure of auditory underst…