{"slug": "duplexgen-adaptive-synthesis-of-human-ai-turn-taking-dialogues", "title": "DuplexGen: Adaptive Synthesis of Human-AI Turn-Taking Dialogues", "summary": "Researchers introduce DuplexGen, a framework that generates human-AI turn-taking dialogues adaptive to specific scenarios by calibrating large language model predictions against a small set of human preference annotations. In six cooperative and competitive tasks, DuplexGen aligned substantially more closely with human turn-taking preferences than uncalibrated prompting or training on generic human-human data, demonstrating that human calibration enables scenario-specific turn-taking synthesis.", "body_md": "arXiv:2607.26178v1 Announce Type: new\nAbstract: Turn-taking is a central component of full-duplex interaction. Which turn-taking behaviors are appropriate varies with the scenario, yet current models apply a single norm regardless of context. This limitation originates in their training data: human-human speech corpora capture natural timing phenomena but provide little role grounding or scenario-specific norms, while heuristic or prompted synthesis methods inject turn-taking behaviors without basing them on human preferences. We introduce DuplexGen, a framework for generating dialogues with scenario-adaptive turn-taking by calibrating LLM predictions against a small set of slot-level human preference annotations. In six cooperative and competitive tasks, human turn-taking preferences differ systematically, and DuplexGen aligns substantially more closely with those preferences than uncalibrated prompting or training solely on generic human-human data; a full-duplex model trained on DuplexGen-generated data exhibits distinctive, human-preferred turn-taking behaviors. These results show that human calibration, not corpus scale or prompt design alone, is what allows turn-taking synthesis to be scenario-specific.", "url": "https://wpnews.pro/news/duplexgen-adaptive-synthesis-of-human-ai-turn-taking-dialogues", "canonical_source": "https://arxiv.org/abs/2607.26178", "published_at": "2026-07-30 04:00:00+00:00", "updated_at": "2026-07-30 04:35:14.511697+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-research", "ai-agents", "natural-language-processing"], "entities": ["DuplexGen", "arXiv"], "alternates": {"html": "https://wpnews.pro/news/duplexgen-adaptive-synthesis-of-human-ai-turn-taking-dialogues", "markdown": "https://wpnews.pro/news/duplexgen-adaptive-synthesis-of-human-ai-turn-taking-dialogues.md", "text": "https://wpnews.pro/news/duplexgen-adaptive-synthesis-of-human-ai-turn-taking-dialogues.txt", "jsonld": "https://wpnews.pro/news/duplexgen-adaptive-synthesis-of-human-ai-turn-taking-dialogues.jsonld"}}