Saved in:
Bibliographic Details
Main Authors: Chang, Kai-Wei, Hu, En-Pei, Kuan, Chun-Yi, Ren, Wenze, Chen, Wei-Chih, Lin, Guan-Ting, Tsao, Yu, Sun, Shao-Hua, Lee, Hung-yi, Glass, James
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2509.26388
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917453279264768
author Chang, Kai-Wei
Hu, En-Pei
Kuan, Chun-Yi
Ren, Wenze
Chen, Wei-Chih
Lin, Guan-Ting
Tsao, Yu
Sun, Shao-Hua
Lee, Hung-yi
Glass, James
author_facet Chang, Kai-Wei
Hu, En-Pei
Kuan, Chun-Yi
Ren, Wenze
Chen, Wei-Chih
Lin, Guan-Ting
Tsao, Yu
Sun, Shao-Hua
Lee, Hung-yi
Glass, James
contents Conversational Spoken Language Models (SLMs) are emerging as a promising paradigm for real-time speech interaction. However, their capacity of temporal dynamics, including the ability to manage timing, tempo and simultaneous speaking, remains a critical and unevaluated challenge for conversational fluency. To address this gap, we introduce the Game-Time Benchmark, a framework to systematically assess these temporal capabilities. Inspired by how humans learn a language through language activities, Game-Time consists of basic instruction-following tasks and advanced tasks with temporal constraints, such as tempo adherence and synchronized responses. Our evaluation of diverse SLM architectures reveals a clear performance disparity: while state-of-the-art models handle basic tasks well, many contemporary systems still struggle with fundamental instruction-following. More critically, nearly all models degrade substantially under temporal constraints, exposing persistent weaknesses in time awareness and full-duplex interaction. The Game-Time Benchmark provides a foundation for guiding future research toward more temporally-aware conversational AI. Demos and datasets are available on our project website https://ga642381.github.io/Game-Time.
format Preprint
id arxiv_https___arxiv_org_abs_2509_26388
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Game-Time: Evaluating Temporal Dynamics in Spoken Language Models
Chang, Kai-Wei
Hu, En-Pei
Kuan, Chun-Yi
Ren, Wenze
Chen, Wei-Chih
Lin, Guan-Ting
Tsao, Yu
Sun, Shao-Hua
Lee, Hung-yi
Glass, James
Audio and Speech Processing
Artificial Intelligence
Computation and Language
Conversational Spoken Language Models (SLMs) are emerging as a promising paradigm for real-time speech interaction. However, their capacity of temporal dynamics, including the ability to manage timing, tempo and simultaneous speaking, remains a critical and unevaluated challenge for conversational fluency. To address this gap, we introduce the Game-Time Benchmark, a framework to systematically assess these temporal capabilities. Inspired by how humans learn a language through language activities, Game-Time consists of basic instruction-following tasks and advanced tasks with temporal constraints, such as tempo adherence and synchronized responses. Our evaluation of diverse SLM architectures reveals a clear performance disparity: while state-of-the-art models handle basic tasks well, many contemporary systems still struggle with fundamental instruction-following. More critically, nearly all models degrade substantially under temporal constraints, exposing persistent weaknesses in time awareness and full-duplex interaction. The Game-Time Benchmark provides a foundation for guiding future research toward more temporally-aware conversational AI. Demos and datasets are available on our project website https://ga642381.github.io/Game-Time.
title Game-Time: Evaluating Temporal Dynamics in Spoken Language Models
topic Audio and Speech Processing
Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2509.26388