Metraipolis › archive › dc_2fcc82ff8db2

When Models Hear What They Expect: Diagnosing Prosodic Heuristics in Multimodal Sarcasm Detection — arXiv:2608.30204

Source: arXiv (en)
Original: https://arxiv.org/abs/2608.30204
License (as archived): CC0-1.0
Archived: 2026-10-17T08:07:25.121Z
SHA-256 of the archived text: 16342b5e4eedb766e6db89d5716eeeeef83f997a90f06d5eb94bcd2ab20012d1
Stable URL of this page: https://metraipolis.com/documento/dc_2fcc82ff8db2

How to cite

When Models Hear What They Expect: Diagnosing Prosodic Heuristics in Multimodal Sarcasm Detection — arXiv:2608.30204 — arXiv (en), https://arxiv.org/abs/2608.30204 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2fcc82ff8db2 (Metraipolis archive dc_2fcc82ff8db2, sha256 16342b5e4eedb766e6db89d5716eeeeef83f997a90f06d5eb94bcd2ab20012d1)

BibTeX

@misc{dc2fcc82ff8db2,
  title = {When Models Hear What They Expect: Diagnosing Prosodic Heuristics in Multimodal Sarcasm Detection — arXiv:2608.30204},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.30204},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2fcc82ff8db2; sha256 16342b5e4eedb766e6db89d5716eeeeef83f997a90f06d5eb94bcd2ab20012d1},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.30204. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Multimodal Large Language Models (MLLMs) process speech and text jointly, yet whether they exploit prosodic cues for pragmatic inference or rely on surface acoustic patterns has received little systematic investigation. We address this through sarcasm detection, evaluating Qwen2.5-Omni and Qwen3-Omni on Mandarin Chinese and English under five modality conditions that decompose the contributions of lexical content, vocal semantics, and prosodic structure. Adding audio systematically inflates false positives without improving true positive detection. Acoustic error diagnosis reveals that model errors cluster on a shared stereotype of expressive prosody, namely elevated pitch and irregular pausing, that diverges from the actual cues marking sarcasm in both languages. Targeted manipulation of only these two dimensions causally confirms the heuristic, inducing false positive rates of up to 60%. Applying the same manipulation template to Gemini~3 Flash Preview without modification replicates the effect, suggesting that the stereotype extends beyond the Qwen Omni family rather than arising from a single model architecture.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-01 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2608.30204 — 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 2026-10-17T07:47:52.787Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).