192815 research outputs found
Sort by
Creating an Effective Methodology for End-User Engagement in AI Auditing
No full textA methodology explains the object of an AI-audit. This object has three loci: identifying significant events (harms or risks), governance (model is behaving as expected), and assurance (trust). The methodology in this paper is being developed as part of the PHAWM project (The Participatory Harm Auditing Workbenches and Methodologies project can be found at https://phawm.org), which seeks to design a workbench that supports inclusive, participant-led auditing of AI application across a range of domains. Project participants range from health service users, parents of school-aged children, to museum professionals and librarians. The project addresses a key gap in existing approaches: the absence of human-centred infrastructures that empower end-users to identify events (An event refers to an occurrence triggered by an AI application that may affect entities and has associated metrics. Each event can be assessed for likelihood, magnitude, and positive or negative valence. We avoid the term harm in our methodology due to its subjectivity, although we acknowledge its common use, including in our own project title, within AI auditing discourse), understand system behavior and participate meaningfully in audit processes
The Future of Geothermal in the United Kingdom: Affordable, Renewable, and Locally Produced Energy for a Resilient Future
Get PDFGeothermal energy can become a cornerstone of the United Kingdom’s
future energy system—yet it is often overlooked. With a growing pipeline of heat projects and a domestic resource for nationwide heating and cooling and selective electricity generation, the UK can mitigate exposure to future external shocks, and strengthen energy security, while creating tens of thousands of jobs, lowering bills, and meeting binding climate targets
Leonard V. Smith, French Colonialism: From the Ancien Régime to the Present (Cambridge University Press, 2023), 249 pp.
Get PDFNo abstract available
Review of Rebecca Helm: How Juries Work: And How They Could Work Better
No full textNo abstract available
Pain fallibility
No full textCall the idea that we are sometimes mistaken about our occurrent pains [italicize] pain fallibility. According to pain fallibility, there is nothing that is plausibly painful about which we are plausibly infallible. While we might assume the truth or falsity of pain's fallibility, direct arguments are thin on the ground. This paper addresses this lacuna and argues for pain fallibility through careful attention to the literature on the philosophy and science of pain, highlighting broad practical implications for addressing pain and theoretical implications for fallibility more generally
Fibered ribbon pretzels
Get PDFWe classify fibered ribbon pretzel knots up to mutation. The classification is complete, except perhaps for members of Lecuona's “exceptional” family of Lecuona [Algebr. Geom. Topol. 15 (2015), no. 4, 2133–2173]. The result is obtained by combining lattice embedding techniques with Gabai's classification of fibered pretzel knots, and exhibiting ribbon disks, some of which lie outside of known patterns for standard pretzel projections
Improved computational models for temperature depression of developed cavitation on hydrofoils
Get PDFTemperature depression in the cavitation zone is necessary for the thermodynamic effect of developed cavitation of a liquid flowing in a device. Proper estimation of temperature depression is useful for determining the required net positive suction head of a rotodynamic pump operating under cavitating flow conditions, where the thermodynamic effect is dominant. In this paper, following a brief review of computational models for temperature depression in the literature, a novel model is identified for temperature depression of cavitating water flowing over hydrofoils. However, this model fails to take account of important variables such as the vapor volume flow rate coefficient and cavity area coefficient and the cavitation number. To remedy this deficiency, two improved computational models are developed by introducing the vapor volume flow rate coefficient and cavity area coefficient and the cavitation number, respectively. The two models are validated by employing experimental data on developed cavitation of water flowing around NACA 0015 hydrofoils at 12° angle of attack. They are best-fitted to experimental curves of temperature depression vs temperature and temperature depression vs cavitation number, respectively, with respective errors of 4.87%–10.91% and 5.11%–8.30%. The determined constants in the improved models are appropriate for prediction of temperature depression–temperature curves. The incorporation of Reynolds number and Mach number slightly reduces the error of temperature depression–temperature curve fitting at both low far-field temperature and large cavitation number
Enhancing data efficiency with a trustworthy counterfactual generative model
Get PDFLeveraging limited data to synthesize an additional training set is essential for robotic vision, particularly in dynamic environments where collecting large datasets is impractical. Traditional robotic vision systems rely on extensive training data for object recognition and scene understanding but struggle to generalize to real-world variations, such as lighting conditions, occlusions, and sensor noise. This article proposes causal diffuse variational autoencoder (causal DiffuseVAE), a novel method integrating causal inference with high-fidelity image synthesis to generate counterfactual images. By combining the disentanglement properties of variational autoencoders (VAEs) with the generative capabilities of diffusion models, causal DiffuseVAE produces realistic, interpretable simulations of variations, such as shadows and occlusions. This combination enables data-efficient generative modeling by learning from small subsets and synthesizing missing or unseen samples. In addition, causal inference ensures that generated data follow real-world dependencies, making it robust and interpretable for deployment in unpredictable environments. Four baseline approaches are evaluated across six different datasets, demonstrating that causal DiffuseVAE consistently outperforms the four baseline approaches
Data sources:
Access Repository Dashboard
Do you manage Open Research Online? Become a CORE Member to access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard! 👇