When Models Hear What They Expect: Diagnosing Prosodic Heuristics in Multimodal Sarcasm Detection
Sarcasm detection is addressed 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 to reveal a shared stereotype of expressive prosody.