Reproduce “PERSIST: Who-What-When Memory Across Sessions for Full-Duplex Spoken Dialogue” 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

Achira Lin, Siyuan Hou, Wenyi Yu, Xinnian Zhao, Haoyu Niu, Wang Geng, Longshuai Xiao, Shihai Xiao and 2 more

PERSIST helps a voice assistant retrieve earlier spoken events by matching what was said with who said it and when. It stores readable, timestamped, speaker-grounded records and combines semantic and temporal scores to rank them. On SpokenTrace, it reaches 85.08% strict answer accuracy with an external writer; this is distinct from retrieval accuracy and does not establish real-world performance.

Summary by alphaXiv.

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