Metraipolis › archive › dc_6d137ed54fdd
Broadening Uncertainty Estimation for Audio Question Answering Across Methods, Formats, and Inputs — arXiv:2609.28879
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.28879
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-08T04:27:10.638Z
- SHA-256 of the archived text
5319215a9eb9eeb4a562db70c0319c33d80ee3fd2b398ffd99c3ea8999c597e0- Stable URL of this page
https://metraipolis.com/documento/dc_6d137ed54fdd
How to cite
Broadening Uncertainty Estimation for Audio Question Answering Across Methods, Formats, and Inputs — arXiv:2609.28879 — arXiv (en), https://arxiv.org/abs/2609.28879 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_6d137ed54fdd (Metraipolis archive dc_6d137ed54fdd, sha256 5319215a9eb9eeb4a562db70c0319c33d80ee3fd2b398ffd99c3ea8999c597e0)
BibTeX
@misc{dc6d137ed54fdd,
title = {Broadening Uncertainty Estimation for Audio Question Answering Across Methods, Formats, and Inputs — arXiv:2609.28879},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.28879},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_6d137ed54fdd; sha256 5319215a9eb9eeb4a562db70c0319c33d80ee3fd2b398ffd99c3ea8999c597e0},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.28879. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Audio-language models can produce confident answers unsupported by the audio, motivating uncertainty estimates that identify unreliable responses. We compare probability-based, sampling-based, self-verification, evidential, and contrastive measures across four open-weight models and five audio QA benchmarks. In multiple-choice evaluation, first-token measures are strongest overall, with top-1 probability achieving a mean AUROC of .740, compared with .708 for ten-sample discrete semantic entropy, while requiring no additional model calls. Across four benchmarks, shifting from multiple-choice to open-ended evaluation lowers mean accuracy from 57.6% to 36.6%, yet uncertainty remains predictive of errors: semantic entropy, maximum token entropy, and semantic agreement achieve mean AUROCs of .697, .694, and .693, respectively. To test whether uncertainty reflects the evidence available to answer the question, we perform input ablations that remove either the audio or the question. Across top-1 confidence, entropy, and sampling-based measures, removing audio reduces error-detection AUROC by .101 on average, compared with .010 when removing the question. Together, these results establish efficient uncertainty baselines and show that uncertainty in audio-language models depends substantially more on available audio evidence than on question text. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-25 in cs.AI). Categorie arXiv: Sound (cs.SD) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2609.28879 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-08T03:50:57.971Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-25?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).