Generative artificial intelligence (GenAI) systems are increasingly used by organizations to deliver information to consumers, patients, students, employees, and citizens. These systems can hallucinate, producing plausible but inaccurate responses. A central question for AI-advised decisions is therefore not only whether users rely on inaccurate information, but whether they recognize that a response may require verification. To answer this question, we review emerging empirical evidence relevant to hallucination detection in goal-directed interactions, with a focus on organization-backed AI advisors. We distinguish three constructs that existing studies often conflate: whether users are skeptical of information presented, whether they verify it (distinguishing attempted from successful verification), and whether the result of verification affects reliance on the information. Across studies examining product search, medical decision-making, content generation, and chatbot-assisted tasks, several patterns emerge. Nearly all studies measure reliance, while variables such as user skepticism and verification of the information are more often targeted by an intervention than measured directly. The cues used to prompt scrutiny of the AI response are predominantly related to the AI output, such as source citations. Whereas general warnings about the risk of hallucinations are among the most deployable interventions for organizations, they show the weakest and most mixed effects in the studies reviewed. More targeted cues about where AI systems tend to fail appear promising, however evidence remains limited. Although the existing literature posits that users may be more likely to scrutinize responses related to particular areas of content, no studies varied the content category, leaving this question open for further research. In future research, measuring skepticism and verification separately from reliance may clarify what current evidence shows, what it only implies, and which questions require further exploration.
To address growing computational demands, energy-efficient hardware technologies such as spintronics and neuromorphic computing have attracted significant interest. In particular, magneto-ionics offers a low-power, non-volatile approach to control magnetic properties, making it particularly suitable for manipulating antiferromagnetic (AFM) materials. In this work, we report magneto-ionic control of exchange bias (EB) in Mn1-xCoxN/Co with a compositionally tunable N & eacute;el temperature, T-N. The high T-N in MnN (> 650 K) typically necessitates high-temperature annealing, which triggers uncontrolled thermally induced ion-motion effects. Addition of Co to MnN reduces T-N, enabling robust EB to be established after field cooling from 400 K, while preserving structural integrity. Importantly, EB can be subsequently tuned by voltage, up to a 30% enhancement observed at 100 K alongside an increase in saturation magnetization (up to approximate to 250 emu cm(-3)). Unlike previous works on similar single-layer nitrides, incorporating an additional ferromagnetic Co layer to form an AFM/ferromagnetic bilayer amplifies the voltage-induced effects. This work highlights the dual role of Co addition to MnN: (i) reducing the thermal requirements for setting EB by lowering T-N, and (ii) enhancing electrical control of EB. These results represent a step forward towards the development of low-power voltage-controlled spintronic devices. (c) 2026 Published by Elsevier Ltd on behalf of The editorial office of Journal of Materials Science & ( http://creativecommons.org/licenses/by/4.0/ )
Solitary fibrous tumor (SFT) is an exceedingly rare mesenchymal neoplasm classified as a soft tissue sarcoma (STS). While historically considered indolent, up to 50
Abstract Cardiovascular–kidney–metabolic (CKM) syndrome represents a continuum of interrelated adiposity, insulin resistance, cardiovascular disease, kidney dysfunction, and metabolic disturbances that evolve across the lifespan. Emerging evidence demonstrates that both biological sex and sociocultural gender significantly shape CKM risk, progression, and clinical expression. CKM syndrome pathogenesis reflects complex multisystem interactions involving adipose tissue dysfunction, neurohormonal activation, inflammatory signaling, and vascular impairment, all of which exhibit important sex-specific patterns. This review examines CKM syndrome from a sex- and gender-informed perspective, highlighting how endogenous and exogenous sex hormones, reproductive transitions, pregnancy-related complications, and dietary exposures shape the long-term CKM syndrome risk. Particular attention is given to the roles of estrogen and testosterone in modulating adipose biology, vascular function, and metabolic regulation. Polycystic ovary syndrome is discussed as a model of androgen excess and multisystem metabolic vulnerability that accelerates CKM features. Finally, we address brain vulnerability within CKM syndrome, emphasizing shared inflammatory, vascular, and neuroendocrine mechanisms linking metabolic dysfunction to cognitive decline and neuropsychiatric disorders. Recognizing these interconnected and sex-specific influences is critical for advancing precision prevention and treatment strategies across the CKM syndrome spectrum. Graphical abstract
One method of electric field control of magnetization is to generate a strain in a piezoelectric layer by applying an electric field to it and transfer this strain to a magnetostrictive nanoscale magnet or to the magnetostrictive soft layer of a magnetic tunnel junction deposited on the piezoelectric, thus controlling its magnetization state through inverse magnetostriction (Villari effect). Such strain-mediated electric field control of magnetization offers an extremely energy-efficient method of writing the magnetic state (similar to 100 aJ/bit). This review discusses the development of computing paradigms based on strain control of magnetism, termed "straintronics" or "magnetic straintronics," and their future potential. Here, we discuss various Boolean memory and logic devices as well as non-Boolean and neuromorphic computing devices proposed and experimentally demonstrated along with their potential advantages and key challenges.
