HiPLEX: Hierarchical Policy Factorization for Full-Duplex Speech Language Models
The paper introduces HiPLEX, a reinforcement learning framework that factorizes a pretrained full-duplex text policy into a control policy deciding when to emit content (choosing among pad, epad and con) and a conditional content policy that picks a token only when con is selected. The authors report that across three Moshi seeds on Full-Duplex-Bench v1, HiPLEX reduces takeover rates during natural user pauses and backchannel opportunities and shortens post-interruption response latency versus GRPO, while keeping comparable judged interruption-response quality, and better matches pooled human turn-timing and backchannel-rate marginals on Moshi and PersonaPlex.
Turn-taking and interruption handling are central pain points for real-time voice agents, and separating timing from content optimization offers a reusable training idea.
Computer Science > Sound
arXiv:2610.07727 (cs)
[Submitted on 6 Oct 2026]
Title:HiPLEX: Hierarchical Policy Factorization for Full Duplex Speech Language Models
Authors:Kyudan Jung, Hyunsin Park, Yoonhyung Lee, Jinhwan Park, Jinhyeok Yang, KiHyun Nam, Jaegul Choo, Jinkyu Lee
View a PDF of the paper titled HiPLEX: Hierarchical Policy Factorization for Full Duplex Speech Language Models, by Kyudan Jung and 7 other authors