{"ID":5438732,"CreatedAt":"2026-07-01T01:17:58.482524686Z","UpdatedAt":"2026-07-03T08:38:05.6384997Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2606.31338","arxiv_id":"2606.31338","title":"Beyond Binary Instrument QA: Probing Instrument Grounding in Music Audio-Language Models","abstract":"Recent music audio-language models achieve high accuracy on instrument question-answering benchmarks, but it remains unclear whether this reflects robust audio grounding or benchmark-specific shortcuts. In this paper, we introduce an OpenMIC-derived diagnostic benchmark sequence for instrument grounding in music audio-language models, extending binary instrument-presence QA to genre-prior-reduced examples, confusable instrument discrimination, longer audio context, and temporal localization. Across these settings, high binary QA accuracy often fails to predict model behavior: models can exhibit option-position bias, confusable-instrument errors, and temporal response bias. These results suggest that instrument grounding should be evaluated with multi-axis diagnostic benchmarks rather than a single aggregate accuracy.","short_abstract":"Recent music audio-language models achieve high accuracy on instrument question-answering benchmarks, but it remains unclear whether this reflects robust audio grounding or benchmark-specific shortcuts. In this paper, we introduce an OpenMIC-derived diagnostic benchmark sequence for instrument grounding in music audio-...","url_abs":"https://arxiv.org/abs/2606.31338","url_pdf":"https://arxiv.org/pdf/2606.31338v1","authors":"[\"Yujun Lee\",\"Joonhyeok Shin\",\"Hyoeun Kim\",\"Kyuhong Shim\"]","published":"2026-06-30T08:39:56Z","proceeding":"cs.SD","tasks":"[\"cs.SD\",\"cs.AI\",\"eess.AS\"]","methods":"[\"Language Model\"]","has_code":false}