324139 research outputs found
Sort by
Measurement of CP violation in B 0 → D + D − and B s 0 → D s + D s − decays
Get PDFA time-dependent, flavour-tagged measurement of CP violation is performed with B0→ D+D− and Bs0→Ds+Ds− decays, using data collected by the LHCb detector in proton-proton collisions at a centre-of-mass energy of 13 TeV corresponding to an integrated luminosity of 6 fb−1. In B0→ D+D− decays the CP-violation parameters are measured to beSD+D−=−0.552±0.100stat±0.010syst, CD+D−=0.128±0.103stat±0.010syst. In Bs0→Ds+Ds− decays the CP-violating parameter formulation in terms of ϕs and |λ| results inϕs=−0.086±0.106stat±0.028systrad, ∣λDs+Ds−∣=1.145±0.126stat±0.031syst. These results represent the most precise single measurement of the CP-violation parameters in their respective channels. For the first time in a single measurement, CP symmetry is observed to be violated in B0→ D+D− decays with a significance exceeding six standard deviations
A simulation study on interactions between defects in diamond
Get PDFThe negatively charged nitrogen-vacancy(NV−) defect in diamond has many applications such as sensing and nanoscale imaging. It is recently becoming a promising
candidate in quantum technologies such as quantum computing and quantum communication. Ultrafast laser fabrication of NV− in diamond has proven its ability to generate NV−with demanding position accuracy with minimal lattice damage. However, there is a possibility that the formed NV− would be destroyed and is assumed to be the consequence of the presence of the nearby interstitial carbon. The GR1 centre in diamond is a common optical centre attributed to the neutral monovacancy of the carbon atom (V0). In this thesis, ab initio simulations are conducted to investigate the ground state electronic interaction between interstitial carbon and neutral monovacancy, interstitial carbon and negatively charged nitrogen-vacancy in diamond. The results suggest that their optical properties might change due to hybridized orbitals within them. Additionally, a carbon-nitrogen machine-learning potential is developed for investigating the diffusion mechanism of interstitial carbon with the presence of NV− centre. According to the molecular dynamics simulations, it could either diffuse away or recombine with the position of vacancy, depending on its initial position and diffusion time. Furthermore, linear-scaling time-dependent density functional theory (TDDFT) is performed, trying to interpret the changes in the optical spectrum during laser annealing where various levels of hybridized orbitals are observed, potentially mapping the difference of photoluminescence signals
From public health to AI safety: improving machine learning approaches by collecting, selecting, or reducing the need for high-quality data
Get PDFIn this thesis, we make advances in several application areas, from public health to AI safety. In the process, we develop novel machine learning (ML) methods to tackle various challenges. In each application area, we take a data-centric approach: we develop a method to automatically select high-quality training data, devise and execute rigorous processes for large-scale manual data collection, generate synthetic training data when data is otherwise hard to obtain, and collate highly diverse test data to evaluate the generalisability of our method.
First, we tackle AI misuse and misalignment by building a detector for “lies”---which we define as particular types of falsehoods---output by large language models (LLMs). This endeavour is complicated by the difficulty in acquiring appropriate training data. In the case of misuse, an attacker might use a wide range of types of LLM or lie-generation methods; obtaining a sufficiently diverse dataset to cover all situations is challenging. In the case of misalignment, the lies can be subtle, such as convincing but false text that plays to human biases; identifying a sufficient number of examples to build a dataset from can be difficult. Our approach offers a potential remedy: We generate training data by simply prompting an LLM with instructions to lie, making it easy to collect a large dataset. We then collate a large, diverse test dataset and extensively study the generalisation of our lie detector to previously unseen settings. We find that it generalises surprisingly well.
In the second part of the thesis, we deal with a situation where we have abundant training data available, but some of it is of low quality. This is common in modern ML, as deep learning models require vast amounts of data, often scraped from the internet. We develop a method to automatically select the most useful data at any given point in training, enabling us to train neural networks in fewer steps, to higher accuracy, and with lower computational costs.
In the third part of the thesis, a necessity for exceptionally high data-quality arises due to the high stakes of the modelling outcomes. During the COVID-19 pandemic, governments worldwide needed information on which non-pharmaceutical interventions (NPIs) could most effectively curb virus transmission. While this question could be tackled with modelling approaches, existing data on the timing and nature of interventions across countries was incomplete and inaccurate. To guide high-stakes policy decisions, we need carefully validated, high-quality data. We collect two datasets with extensive quality-control measures as the basis of our modelling studies. We then fit multiple Bayesian models to infer the effect of various government interventions on viral transmission. Our effectiveness estimates are backed with rigorous model validation experiments, a step often lacking in other work on this topic. The resulting NPI effectiveness estimates informed policy decisions in various countries.
Overall, we make progress in different application areas by leveraging different data-centric approaches: generating synthetic training data, collating diverse test data to evaluate our method, selecting the most useful data when data is abundant, and collecting high-quality data when existing datasets are insufficient
Perceptions of meteorological phenomena in ancient Egypt
Get PDFPerception of landscape has become crucial for how we recreate ancient experience, linking landscape to concepts of embodiment, identity, and dwelling, but the perception of weather has been absent from these discussions. While historical climatology has reconstructed past climates using quantitative methods, limited research has been done on the impact of weather phenomena, both mundane and extraordinary, on cultural experience and expression. This dissertation applies landscape theory to the weather of ancient Egypt, studying how ancient Egyptians perceived, conceptualised, and interacted with their weather, through archaeological, representational, and textual sources.
The first chapter discusses the anthropological approach to weather as part of a culture’s landscape, which forms the theoretical backbone of this dissertation. The following chapter provides an outline of Egyptian climate and weather patterns, creating a baseline from which to build when discussing ancient perception and conceptualisation. My aim is not to reconstruct the palaeoclimate, but to define the meteoric possibilities. The third chapter examines the archaeological and representation evidence for engagement with the weather, whilst the fourth chapter explores the lexicon of weather, refining our understanding of these terms by adding nuance. Finally, the discussion brings together all the evidence in order to appraise the perception of weather in ancient Egypt holistically.
The dissertation has a broad scope, considering evidence from the Old Kingdom to the end of the New Kingdom, in order to maximise the potential of the fragmentary evidence. These varied sources indicate the practical, religious, and cultural relationships the Egyptians had with their weather: they built structures to both utilise and mitigate against weather, incorporated weather into cosmogonies and their conceptualisation of the ongoing maintenance of the universe, and used weather terminology in metaphorical language, among other aspects. I argue that the Egyptians were deeply entwined with their weather, just as they were with the rest of their landscape
Understanding the private-public school performance gap in PISA: evidence from Portugal
Get PDFWe analyse the PISA-reported convergence in the performance of private and public schools in Portugal. When PISA sampling weights are used, the number of students enrolled in those types of schools and specific grades/tracks of study differs significantly from official population figures. To account for those differences, we apply a post-stratification adjustment; however, sample sizes are small, resulting in estimates with low precision for several subgroups. We propose recommendations for improving the handling of these issues in future PISA cycles. In an additional analysis, we also account for changes in the composition of the student population. When all factors are considered, the convergence in scores is far less impressive than reported. For instance, in Science, after adjusting the sampling weights and removing population composition effects, the reported convergence of 46 points between private and public schools from 2015 to 2018 amounts to only 9 points. The decomposition and sample adjustment methods used in this paper can be easily adapted to other contexts
Amplitude analysis of B + → ψ (2 S ) K + π + π − decays
Get PDFThe first full amplitude analysis of B+ → ψ(2S)K+π+π− decays is performed using proton-proton collision data corresponding to an integrated luminosity of 9 fb−1 recorded with the LHCb detector. The rich K+π+π− spectrum is studied and the branching fractions of the resonant substructure associated with the prominent K1(1270)+ contribution are measured. The data cannot be described by conventional strange and charmonium resonances only. An amplitude model with 53 components is developed comprising 11 hidden-charm exotic hadrons. New production mechanisms for charged charmonium-like states are observed. Significant resonant activity with spin-parity JP = 1+ in the ψ(2S)π+ system is confirmed and a multi-pole structure is demonstrated. The spectral decomposition of the ψ(2S)π+π− invariant-mass structure, dominated by X0 → ψ(2S)ρ(770)0 decays, broadly resembles the J/ψϕ spectrum observed in B+ → J/ψϕK+ decays. Exotic ψ(2S)K+π− resonances are observed for the first time
Decentralized convergence to equilibrium prices in trading networks
Get PDFWe propose a decentralized market model in which agents
can negotiate bilateral contracts. This builds on a similar,
but centralized, model of trading networks introduced by
Hatfield et al. (2013). Prior work has established that fullysubstitutable preferences guarantee the existence of competitive equilibria which can be centrally computed. Our motivation comes from the fact that prices in markets such as
over-the-counter markets and used car markets arise from decentralized negotiation among agents, which has left open
an important question as to whether equilibrium prices can
emerge from agent-to-agent bilateral negotiations. We design
a best response dynamic intended to capture such negotiations between market participants. We assume fully substitutable preferences for market participants. In this setting,
we provide proofs of convergence for sparse markets (covering many real world markets of interest), and experimental results for more general cases, demonstrating that prices
indeed reach equilibrium, quickly, via bilateral negotiations.
Our best response dynamic, and its convergence behavior,
forms an important first step in understanding how decentralized markets reach, and retain, equilibrium
Hydrogen borrowing-based methods for the construction of quaternary stereocentres
Get PDFCompounds containing quaternary stereocentres are a valuable motif in biologically active compounds. Herein we present our strategy to utilise the hydrogen borrowing manifold to access α-quaternary ketones via a tandem acceptorless dehydrogenation-cyclisation cascade. This new application of the methodology results in the formation of five- and six-membered carbocycles with a high degree of diastereoselectivity. Interestingly, benzylic alcohol substrates behaved anomalously and eliminated sulfinate in situ to give a set of rearranged α-quaternary ketone products
Conflict abroad and political trust at home evidence from a natural experiment
Get PDFDo conflicts abroad affect trust at home? While we know that conflicts impact trust in warring
countries, we lack evidence on whether people in neighbouring, but non-involved, countries are also
affected. We address this question in the case of Russia’s invasion of Ukraine in February 2022, which
represented a large shock to the security and economy of European countries. Our identification strategy uses the overlap between the timing of the Russian invasion and the European Social Survey fieldwork in eleven European countries. We find that the invasion increased respondents’ trust in their
country’s politicians, political parties, and national parliaments, as well as satisfaction with the government. Further analyses using other surveys and previous conflicts suggest this effect depends on
proximity to the conflict and the political regimes of the attacked country. These findings contribute
to our understanding of the complex and indirect effects of conflicts on domestic political trust
Friend request accepted: fundamental features of social environments determine social affiliation decisions
Get PDFHumans start new friendships and social connections throughout their lives, and it has been consistently found that such relationships lead to mental and physical well-being. In this thesis, I investigated the behavioural and neural mecha- nisms governing our decisions to initiate friendships with other people. I examined whether such decisions are influenced by the friendliness, i.e., the social reward rate, and the density, i.e., the rate of opportunities, afforded by the environment. In a computer based online task (n=783), I found that people were more likely to send friend requests in friendly and sparse environments in comparison to hostile and dense environments. Further, I found task-related measures like overall friend requests were correlated with personality-related factors like social thriving. Next, in a 7T fMRI study (n=24), I found that the subcortical dorsal raphe nucleus represented density-related effects and the substantia nigra (SN) represented friendliness-related effects. Further, cortical regions like the anterior insula (aI) represented both friend- liness and density related effects. Next, in resting state fMRI data (n=400), I showed that model predicted factor score corresponding to anhedonia was related to functional connectivity between the SN and the aI. Finally, in a social learning task, I found that people took background statistics of an environment into account when deciding between learning from others and foraging for food by themselves. Taken together, these findings suggest that the human brain takes background statistics of an environment into account while making social decisions and that such decisions can be explained by personality or psychiatric factors