Sep 2026· International Journal of Human-Computer Interaction· pp. 1-14
Safety Warnings and Signage
Abstract
This study examined whether Generative Artificial Intelligence (AI) can evaluate in-vehicle drunk-driving warning symbols (telltales) in ways comparable to human perceptual and semantic judgments. Using visibility and intuitiveness ratings collected from Korean and Texas-based U.S. participants, the same 30 telltales were evaluated by GPT-4o and GPT-5.5. Pearson correlation and agreement analyses revealed significant positive associations between AI and human evaluations for both metrics, with stronger correspondence in intuitiveness. Symbols with clear meanings, simple structures, and familiar alcohol-, vehicle-, or prohibition-related cues received consistently favorable ratings from both AI and humans. These results suggest that AI judgments may be associated with symbolic meaning as well as visual form, supporting its potential as a preliminary evaluation tool. In early design stages of safety-critical domains such as drunk-driving prevention, AI may help prescreen visual designs, improve evaluation efficiency, strengthen preliminary design review, and support human-centered telltale development for practical vehicle interfaces.
GAOKAO-Bench is introduced, an intuitive benchmark that employs questions from the Chinese GAOKAO examination as test samples, including both subjective and objective questions that contribute a robust evaluation benchmark for future large language models and offers valuable insights into the advantages and limitations...
Xiaotian Zhang, Chun-yan Li, Yi Zong et al.· arXiv.org· 216 citations· ⚡17
This work investigates the possibilities of using LLMs in a resume screening setting via a document retrieval framework that simulates job candidate selection and finds that the MTEs are biased, significantly favoring White-associated names in 85% of cases and female-associated names in only 11.1% of cases.
This work shows that orders of magnitude enhancement in performance could be obtained by a combination of hardware improvements and tight quantum-HPC integration and introduces high-performance architectures for quantum-probabilistic computing with custom-designed accelerators to tackle today's industry-scale classical...
Masoud Mohseni, Artur Scherer, K. Johnson et al.· arXiv.org· 121 citations· ⚡9
This paper presents a comprehensive overview of the Ultralytics YOLO family, emphasizing architectural evolution, benchmarking, deployment, and emerging directions from YOLOv5 through YOLO27, and examines detection, segmentation, depth, classification, pose, oriented detection, tracking, export, quantization, and deplo...
This work revisits schema linking when using the latest generation of large language models (LLMs) and finds empirically that newer models are adept at utilizing relevant schema elements during generation even in the presence of large numbers of irrelevant ones.
Karime Maamari, Fadhil Abubaker, Daniel Jaroslawicz et al.· arXiv.org· 109 citations· ⚡19
A novel threat is unveiled in which attackers steer the RAG system's response by injecting malicious passages into its knowledge base, enabling the attacker to steer the response without altering the user input or modifying the RAG weights.
Jiaqi Xue, Meng Zheng, Yebowen Hu et al.· arXiv.org· 109 citations· ⚡8
Related blog posts
MIT News · Artificial Intelligence· news.mit.eduSep 30, 2026
With $2.1 million funding from Google.org, the open-source Public Transit Intelligence Hub will unify public transit monitoring, operations, and passenger communication.
