@misc{indiciae708ec5bba893,
  title = {D2: High-Capability Full-Duplex Spoken Language Models from Half-Duplex Backbones},
  author = {Cheng-Kuang Chang and Alexander H. Liu and Brian Kingsbury and Hilde Kuehne and Rogerio Feris and Samuel Thomas and James Glass and Kai-Wei Chang},
  year = {2026},
  url = {https://arxiv.org/abs/2610.17124},
  note = {Source identifier: 2610.17124}
}