Metraipolis › archive › dc_deda2e4cb9ad
XTurnix: Large-Scale Self-Supervised Turn Control through Two-State Binary Decisions — arXiv:2610.04400
- Source
- arXiv
- Original
- https://arxiv.org/abs/2610.04400
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-06T05:26:11.654Z
- SHA-256 of the archived text
3992ebf9a7b48cd2c8380902a8ea20a6976ead8a402f38341734330f80f8c280- Stable URL of this page
https://metraipolis.com/documento/dc_deda2e4cb9ad
How to cite
XTurnix: Large-Scale Self-Supervised Turn Control through Two-State Binary Decisions — arXiv:2610.04400 — arXiv, https://arxiv.org/abs/2610.04400 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_deda2e4cb9ad (Metraipolis archive dc_deda2e4cb9ad, sha256 3992ebf9a7b48cd2c8380902a8ea20a6976ead8a402f38341734330f80f8c280)
BibTeX
@misc{dcdeda2e4cb9ad,
title = {XTurnix: Large-Scale Self-Supervised Turn Control through Two-State Binary Decisions — arXiv:2610.04400},
howpublished = {arXiv},
url = {https://arxiv.org/abs/2610.04400},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_deda2e4cb9ad; sha256 3992ebf9a7b48cd2c8380902a8ea20a6976ead8a402f38341734330f80f8c280},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.04400. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
General turn-taking behavior in real-time dialogue systems requires deciding whether to keep listening or start responding while listening, and whether to continue or stop while speaking. Existing turn detectors use heterogeneous, task-specific label spaces and are often trained on limited annotations or evaluated on isolated utterances, making them difficult to use as a unified causal controller with comprehensive context. We propose XTurnix, a compact text-based model that formulates turn control as two binary decisions conditioned on the AI's current listening or speaking state and predicts a single control token from the complete dialogue history. XTurnix is pretrained on 5.5 million causal action examples automatically derived from timestamped two-speaker transcripts, then fine-tuned on synthetic multi-turn examples with a flatter distribution across the four state-action labels. We evaluate XTurnix on four public benchmarks and a balanced self-curated benchmark. Across the public benchmarks, XTurnix achieves the best results on all SemanticVAD and LiveKit splits, ties the native Smart-Turn model on Smart-Turn Bench, and achieves the highest incomplete-turn accuracy on Easy-Turn. On the self-curated benchmark, it reaches 89.06% accuracy, more than 20 percentage points above the strongest third-party baseline at 68.75%, while maintaining F1 scores between 84.21% and 90.63% across all four categories. These results demonstrate unified listening- and speaking-state turn control in a single compact model. Code is available at this https URL , with an interactive demo at this https URL . Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC). Pagina del lavoro: https://arxiv.org/abs/2610.04400 — 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 2026-10-06T05:24:05.846Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).