Multi-Faceted Interactivity Alignment in Full-Duplex Speech Models

arXiv:2606.11167 · cs.CL, eess.AS · Submitted 2026-06-09 · Read on arXiv

cs.CL, eess.AS

Submitted: 2026-06-09

Updated: 2026-08-28

Comments: Accepted to EMNLP 2026 Main Conference

Code: https://github.com/NVIDIA/personaplexhttps:

License: http://creativecommons.org/licenses/by-nc-sa/4.0/

The gist: Full-duplex spoken dialogue models can listen and speak simultaneously, making them a promising architecture for natural conversation.

Terminology

Abstract

Full-duplex spoken dialogue models can listen and speak simultaneously, making them a promising architecture for natural conversation. However, current models are trained solely with supervised learning through token-level likelihood maximization, which does not directly optimize interaction-level behaviors, causing interactivity issues such as excessive silence and ill-timed turn-taking. Recent work has applied reinforcement learning (RL) to improve interactivity, but existing methods address only a limited set of interactive behaviors in their rewards. In this work, we propose a post-training alignment method that comprehensively improves the interactivity of full-duplex spoken dialogue models through RL. We address the four canonical axes of interactivity: pause handling, turn-taking, backchanneling, and user interruption. For each axis, we extract short audio segments from human conversation corpora and optimize the model with axis-specific reward functions. An extra LLM-based reward for response quality prevents semantic degradation. We apply our method to two open-source models, Moshi and PersonaPlex, demonstrating consistent improvements in interactivity on both offline evaluation with pre-recorded audio and real-time multi-turn dialogue evaluation.

Sources

Related papers