arXiv · cs.AI · Liang-Yuan Wu, Zih-Ching Chen, Tongshuang Wu, Chao-Han Huck Yang, Hua Shen · 11 小时前模型评分50
SpeechEQ:评估社交语音对话模型情商的新基准
SpeechEQ: Benchmarking Emotional Intelligence Quotient in Socially Aware Voice Conversational Models
导读
研究者提出 SpeechEQ 基准,用 2,265 段对话、覆盖 15 个情商子量表,评估语音语言模型(SLM,Speech-Language Model)在多轮对话中的社会语言推理能力,并给出 Spoken EQ(SEQ)分数。实验显示端到端模型优于级联系统,但仍受制于依赖文本的“模态捷径”、对齐引发的“安全陷阱”和“上下文失忆”。该基准已公开数据集与演示页面。
分类置信度
中 · 待人工复核
- 只有单一媒体来源,需要人工复核。
来源级别:媒体与个人
事件类型:研究发布
证据摘录
- “We introduce \textsc{SpeechEQ}, a comprehensive framework designed to evaluate the sociolinguistic reasoning of Speech-Language Models (SLMs).”
- “The framework includes a validated dataset of 2,265 dialogues across 15 Emotional Quotient (EQ) subscales grounded in EQ-i 2.0 theory, along with a multi-turn evaluation protocol measured by our proposed Spoken EQ (SEQ) score inspired by human EQ assessments.”
来源:arXiv · cs.AI · arxiv.org