ACL - arXivDaily 期刊&会议

2512.01109 2025-12-02 cs.CL

How do we measure privacy in text? A survey of text anonymization metrics

如何测量文本中的隐私？文本匿名化度量的调查

Yaxuan Ren, Krithika Ramesh, Yaxing Yao, Anjalie Field

机构 * Johns Hopkins University（约翰霍普金斯大学）

AI总结本文通过系统调查澄清文本隐私度量标准，分析六个隐私概念与法律标准的契合度，为隐私评估提供指导。

Comments 13 pages, 1 figure, 1 table. To be published in Findings of the Association for Computational Linguistics (AACL-IJCNLP 2025). Related resources at: https://github.com/ryxGuo/privacy-metrics-survey

2507.19362 2025-12-02 cs.CV cs.AI cs.CL cs.CY cs.LG

LOTUS: A Leaderboard for Detailed Image Captioning from Quality to Societal Bias and User Preferences

LOTUS: 一种用于从质量到社会偏见和用户偏好的详细图像描述的排行榜

Yusuke Hirota, Boyi Li, Ryo Hachiuma, Yueh-Hua Wu, Boris Ivanovic, Yuta Nakashima, Marco Pavone, Yejin Choi, Yu-Chiang Frank Wang, Chao-Han Huck Yang

机构 * NVIDIA Research（NVIDIA研究）

AI总结 LOTUS是一种用于评估详细图像描述质量、偏见和社会偏见的排行榜，通过定制标准满足不同用户偏好，揭示了模型在不同评估标准上的表现差异。

Comments Accepted to ACL 2025. Leaderboard: huggingface.co/spaces/nvidia/lotus-vlm-bias-leaderboard

2507.19361 2025-12-02 cs.CL cs.AI cs.SC cs.SD eess.AS

SpeechIQ: Speech-Agentic Intelligence Quotient Across Cognitive Levels in Voice Understanding by Large Language Models

SpeechIQ: 大型语言模型在语音理解中的跨认知层级语音智能商

Zhen Wan, Chao-Han Huck Yang, Yahan Yu, Jinchuan Tian, Sheng Li, Ke Hu, Zhehuai Chen, Shinji Watanabe, Fei Cheng, Chenhui Chu, Sadao Kurohashi

机构 * Kyoto University（京都大学）； NVIDIA ； Carnegie Mellon University（卡内基梅隆大学）； Institute of Science Tokyo（东京科学研究院）

AI总结 SpeechIQ是一种基于语音的智能评估框架，通过三个认知层级评估大型语言模型在语音理解中的能力，提供统一的比较和识别标注错误与幻觉。

Comments ACL 2025 main. Our Speech-IQ leaderboard is hosted at huggingface.co/spaces/nvidia/Speech-IQ-leaderboard. Speech-IQ Calculator: https://github.com/YukinoWan/SpeechIQ

2411.05945 2025-12-02 cs.CL cs.AI cs.LG cs.MA eess.AS

NeKo: Cross-Modality Post-Recognition Error Correction with Tasks-Guided Mixture-of-Experts Language Model

NeKo：跨模态识别后纠错与任务引导的专家混合语言模型

Yen-Ting Lin, Zhehuai Chen, Piotr Zelasko, Zhen Wan, Xuesong Yang, Zih-Ching Chen, Krishna C Puvvada, Szu-Wei Fu, Ke Hu, Jun Wei Chiu, Jagadeesh Balam, Boris Ginsburg, Yu-Chiang Frank Wang, Chao-Han Huck Yang

机构 * NVIDIA

AI总结 NeKo通过任务引导的专家混合模型，实现跨模态识别后的高效纠错，显著降低WER并提升BLEU分数。

Comments ACL 2025 Industry Track. NeKo LMs: https://huggingface.co/nvidia/NeKo-v0-post-correction