Reproduce “SALMONN-duo: Adaptive Dual-System Coordination for Full-Duplex Voice Agents” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Wenyi Yu, Siyin Wang, Terumi Chiba, Xianzhao Chen, Xiaohai Tian, Jun Zhang, Lu Lu, Chao Zhang

SALMONN-duo lets a full-duplex speech model continue a conversation while an asynchronous backend handles requests that need knowledge or tools. The frontend decides when to delegate, then incorporates the backend’s result into the ongoing exchange. On TriviaQA, the 8B system reports 80.2% accuracy with 25.0% delegation, versus MoshiRAG’s 75.8% and 99.9%; this compares systems, not isolated routing policies.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.