Determining whether AI systems process information similarly to humans is central to cognitive science and trustworthy AI. While modern AI models can match human accuracy on standard tasks, such parity does not guarantee that their underlying decision-making strategies resemble those of humans. Assessing performance using error alignment metrics to compare how humans and models fail, and how this changes for distorted, or otherwise more challenging, stimuli, provides a viable pathway toward a finer characterization of model-human alignment. However, existing out-of-distribution (OOD) analyses for challenging stimuli are limited due to methodological choices: they define OOD shift relative to model training data or use arbitrary distortion-specific parameters with little correspondence to human perception, hindering principled comparisons. We propose a human-centred framework that redefines the degree of OOD as a spectrum of human perceptual difficulty. By quantifying how much a collection of stimuli deviates from an undistorted reference set based on human accuracy, we construct an OOD spectrum and identify four distinct regimes of perceptual challenge. This approach enables principled model-human comparisons at calibrated difficulty levels. We apply this framework to object recognition and reveal unique, regime-dependent model-human alignment rankings and profiles across deep learning architectures. Vision-language models are most consistently human aligned across near- and far-OOD conditions, but convolutional neural networks (CNNs) are more aligned than vision transformers (ViTs) for near-OOD and ViTs are more aligned than CNNs for far-OOD. Our work demonstrates the critical importance of accounting for cross-condition differences, such as perceptual difficulty, for a principled assessment of model-human alignment.
The Reward Model Selection Crisis in Personalized Alignment
School of Informatics, University of Edinburgh(信息学院,爱丁堡大学);Agency for Science, Technology and Research (A*STAR)(科技研究局);Samsung AI Research Center(三星人工智能研究中心)
LLM Safety Alignment in Low-Resource Languages: A Systematic Literature Review
African Institute for Mathematical Sciences (AIMS)(非洲数学科学研究所);Bayero University Kano(卡诺贝罗大学);Brown University(布朗大学);Centrum Wiskunde & Informatica(数学与计算机科学中心);University of Pretoria(比勒陀利亚大学);University of Hamburg(汉堡大学);Imperial College London(伦敦帝国学院)
Integrating Feature Understandability into Feature Selection for Human-Centred Tabular Explanations
School of Computer Science and Information Technology, University College Cork(科克大学计算机科学与信息技术学院);Centre for Research Training in Artificial Intelligence, University College Cork(科克大学人工智能研究培训中心);Department of Methodology and Statistics, Tilburg University(蒂尔堡大学方法论与统计系);Insight Centre for Data Analytics, University College Cork(科克大学Insight数据分析中心)
Refusal Falls off a Cliff: How Safety Alignment Fails in Reasoning?
Zhejiang University(浙江大学);Hong Kong Polytechnic University(香港理工大学);East China Normal University(华东师范大学);Southern University of Science and Technology(南方科技大学);Renmin University(中国人民大学);Nanyang Technological University(南洋理工大学);Xiaohongshu Inc.(小红书公司);INSAIT
When Alignment Fails: Multimodal Adversarial Attacks on Vision-Language-Action Models
TGAI Lab, School of Engineering, Westlake University(西拉库大学工程学院TGAI实验室);Pennsylvania State University(宾夕法尼亚州立大学);Sony Research, Sony(索尼研究实验室,索尼);Xidian University(西安电子科技大学)
Speech-Text Dialog Pre-training for Spoken Dialog Understanding with Explicit Cross-Modal Alignment
Alibaba Group(阿里巴巴集团);Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences(中国科学院深圳先进技术研究院);University of Chinese Academy of Sciences(中国科学院大学);University of Science and Technology of China(中国科学技术大学)
Positive Alignment: Artificial Intelligence for Human Flourishing
Department of Psychiatry, University of Oxford(牛津大学精神病学系);Flourishing Intelligence Program, Centre for Eudaimonia and Human Flourishing, Linacre College, University of Oxford(牛津大学幸福智能计划、幸福与人类繁荣中心、林acre学院);Google DeepMind(谷歌DeepMind);LIFE;OpenAI;Anthropic;University of California, Los Angeles(加州大学洛杉矶分校);Aily Labs(Aily实验室);Stanford University(斯坦福大学);Tufts University(塔夫茨大学);Positive AI Labs(积极AI实验室);Department of Informatics, University of Sussex(Sussex大学信息学系);Department of Brain Sciences, Imperial College London(伦敦帝国理工学院脑科学系)