arXiv subjects

Anthony Payne

Publications and source records attributed to Anthony Payne.

3 recordsLinked to original sources

arXiv2026

Evaluating Language Models for Harmful Manipulation

Interest in the concept of AI-driven harmful manipulation is growing, yet current approaches to evaluating it are limited. This paper introduces a framework for evaluating harmful AI manipulation via context-specific human-AI interaction studies. We illustrate the utility of this framework by assessing an AI model with 10,101 participants spanning interactions in three AI use domains (public policy, finance, and health) and three locales (US, UK, and India). Overall, we find that that the tested model can produce manipulative behaviours when prompted to do so and, in experimental settings, is able to induce belief and behaviour changes in study participants. We further find that context matters: AI manipulation differs between domains, suggesting that it needs to be evaluated in the high-stakes context(s) in which an AI system is likely to be used. We also identify significant differences across our tested geographies, suggesting that AI manipulation results from one geographic region may not generalise to others. Finally, we find that the frequency of manipulative behaviours (propensity) of an AI model is not consistently predictive of the likelihood of manipulative success (efficacy), underscoring the importance of studying these dimensions separately. To facilitate adoption of our evaluation framework, we detail our testing protocols and make relevant materials publicly available. We conclude by discussing open challenges in evaluating harmful manipulation by AI models.

cs.AI↗

arXiv2025

Single-Molecule Water Motion on h-BN and Graphene: A Paradigm Shift in Understanding the Behaviour of Water on 2D Material Interfaces

Understanding water behaviour on 2D materials is crucial for sensing, microfluidics, and tribology. While water/graphene interactions are well studied, water on hexagonal boron nitride (h-BN) remains largely unexplored. Despite structural similarity to graphene, h-BN's slightly polar B-N bonds impart a large band gap, high thermal conductivity, and chemical stability, making it promising for electronics, lubricants, and coatings. Moreover, existing water studies often focus on multilayer water dynamics, overlooking single-molecular details. We bridge this gap by studying single-molecular water friction and diffusion on h-BN, comparing it with graphene using helium spin-echo experiments and ab initio calculations. Our findings show that water diffusion on h-BN/Ni follows a complex rotational-translational dynamic, unlike graphene. While conventional views treat water motion as discrete jumps between equivalent adsorption sites, we demonstrate that on h-BN, water molecules rotate freely around their centre of mass. Although the binding energies of water on h-BN and graphene are similar, the activation energy for water dynamics on h-BN is 2.5 times lower than on graphene, implying a much lower barrier for molecular mobility. The fundamentally different diffusion characteristics which classical models cannot capture, underscores the need to rethink how we model water on polar 2D materials. Moreover, our analysis reveals that the metal substrate strongly influences water friction, with h-BN/Ni showing a markedly lower friction than graphene/Ni, in stark contrast to the free-standing materials. These findings challenge assumptions about 2D material-water interactions, highlighting the crucial role of substrate effects in chemistry and material science and offer insights for designing next-generation microfluidic devices that require precise water mobility control.

cond-mat.mes-hall↗

arXiv2022

Evolution of ordered nanoporous phases during h-BN growth: Controlling the route from gas-phase precursor to 2D material by $\textit{in-situ}$ monitoring

Large-area single-crystal monolayers of two-dimensional (2D) materials such as graphene and hexagonal boron nitride (h-BN) can be grown by chemical vapour deposition (CVD). However, the high temperatures and fast timescales at which the conversion from a gas-phase precursor to the 2D material appear, make it extremely challenging to simultaneously follow the atomic arrangements. We utilise helium atom scattering to discover and control the growth of novel 2D h-BN nanoporous phases during the CVD process. We find that prior to the formation of h-BN from the gas-phase precursor, a metastable $(3\times3)$ structure is formed, and that excess deposition on the resulting 2D h-BN leads to the emergence of a $(3\times4)$ structure. We illustrate that these nanoporous structures are produced by partial dehydrogenation and polymerisation of the borazine precursor upon adsorption. These steps are largely unexplored during the synthesis of 2D materials and we unveil the rich phases during CVD growth. Our results provide significant foundations for 2D materials engineering in CVD, by adjusting or carefully controlling the growth conditions and thus exploiting these intermediate structures for the synthesis of covalent self-assembled 2D networks.

cond-mat.mtrl-sci↗