Large LMs such as GPT-3 are powerful, but can commit mistakes that are obvious to humans. For example, GPT-3 would mistakenly interpret "What word is similar to good?" to mean a homophone, while the user intended a synonym. Our goal is to effectively correct such errors via user interactions with the system but without retraining, which will be prohibitively costly. We pair GPT-3 with a growing memory of recorded cases where the model misunderstood the user's intents, along with user feedback for clarification. Such a memory allows our system to produce enhanced prompts for any new query based on the user feedback for error correction on similar cases in the past. On four tasks (two lexical tasks, two advanced ethical reasoning tasks), we show how a (simulated) user can interactively teach a deployed GPT-3, substantially increasing its accuracy over the queries with different kinds of misunderstandings by the GPT-3. Our approach is a step towards the low-cost utility enhancement for very large pre-trained LMs. Code, data, and instructions to implement MEMPROMPT for a new task at https://www.memprompt.com/.

The goal of the present paper is to develop and validate a questionnaire to assess AI literacy. In particular, the questionnaire should be deeply grounded in the existing literature on AI literacy, should be modular (i.e., including different facets that can be used independently of each other) to be flexibly applicable in professional life depending on the goals and use cases, and should meet psychological requirements and thus includes further psychological competencies in addition to the typical facets of AIL. We derived 60 items to represent different facets of AI Literacy according to Ng and colleagues conceptualisation of AI literacy and additional 12 items to represent psychological competencies such as problem solving, learning, and emotion regulation in regard to AI. For this purpose, data were collected online from 300 German-speaking adults. The items were tested for factorial structure in confirmatory factor analyses. The result is a measurement instrument that measures AI literacy with the facets Use & apply AI, Understand AI, Detect AI, and AI Ethics and the ability to Create AI as a separate construct, and AI Self-efficacy in learning and problem solving and AI Self-management. This study contributes to the research on AI literacy by providing a measurement instrument relying on profound competency models. In addition, higher-order psychological competencies are included that are particularly important in the context of pervasive change through AI systems.

Lebanon, Slovenia, and the United Arab Emirates (UAE) are the top three countries that are most interested in Artificial Intelligence (AI) crypto, according to CoinGecko's recent report. Countries with major economic problems, like Nigeria, Sri Lanka, and Pakistan, have also ranked higher in the charts -- while the U.S. was placed 33rd, the CoinGecko report stated. The report measured the search popularity of 14 English search terms related to AI crypto between Nov. 30, 2022, and Feb. 16. A 100 indicates maximum popularity, while 50 indicates half -- zero would mean there was not enough data to examine. Lebanon scored 100 on almost all 14 search terms -- collecting 1,200 points and ranking first on the list.

During a trip home to Johannesburg, South Africa, while completing an engineering master's program in Japan, Pelonomi Moiloa attended the largest machine learning community gathering she'd ever seen in Africa, just a few miles from where she grew up. In all, 600 people from 22 nations attended 2017's Deep Learning Indaba, held at the University of Witwatersrand, discussing topics like health care and agriculture solutions custom-made to meet the needs of African people. That week-long gathering made Moiloa feel she could have an impact on the lives of Africans, and it helped convince her to move back to South Africa and look for a way to put her engineering skills to work on her home continent. "The conversations were around making a genuine impact and positive change in African lives on a mass scale, and that was something I really wanted to be a part of," she says. This month, Moiloa will join some organizers of Deep Learning Indaba to launch Lelapa, a commercial and industrial AI research company focused on serving the needs of the 1 billion people in Africa.

Recognizing the country where a picture has been taken has many potential applications, such as identification of fake news and prevention of disinformation campaigns. Previous works focused on the estimation of the geo-coordinates where a picture has been taken. Yet, recognizing in which country an image was taken could be more critical, from a semantic and forensic point of view, than estimating its spatial coordinates. In the above framework, this paper provides two contributions. First, we introduce the VIPPGeo dataset, containing 3.8 million geo-tagged images. Secondly, we used the dataset to train a model casting the country recognition problem as a classification problem. The experiments show that our model provides better results than the current state of the art. Notably, we found that asking the network to identify the country provides better results than estimating the geo-coordinates and then tracing them back to the country where the picture was taken.

When extracting structured data from repetitively organized documents, such as dictionaries, directories, or even newspapers, a key challenge is to correctly segment what constitutes the basic text regions for the target database. Traditionally, such a problem was tackled as part of the layout analysis and was mostly based on visual clues for dividing (top-down) approaches. Some agglomerating (bottom-up) approaches started to consider textual information to link similar contents, but they required a proper over-segmentation of fine-grained units. In this work, we propose a new pragmatic approach whose efficiency is demonstrated on 19th century French Trade Directories. We propose to consider two sub-problems: coarse layout detection (text columns and reading order), which is assumed to be effective and not detailed here, and a fine-grained entry separation stage for which we propose to adapt a state-of-the-art Named Entity Recognition (NER) approach. By injecting special visual tokens, coding, for instance, indentation or breaks, into the token stream of the language model used for NER purpose, we can leverage both textual and visual knowledge simultaneously. Code, data, results and models are available at https://github.com/soduco/paper-entryseg-icdar23-code, https://huggingface.co/HueyNemud/ (icdar23-entrydetector* variants)

Population synthesis consists of generating synthetic but realistic representations of a target population of micro-agents for the purpose of behavioral modeling and simulation. We introduce a new framework based on copulas to generate synthetic data for a target population of which only the empirical marginal distributions are known by using a sample from another population sharing similar marginal dependencies. This makes it possible to include a spatial component in the generation of population synthesis and to combine various sources of information to obtain more realistic population generators. Specifically, we normalize the data and treat them as realizations of a given copula, and train a generative model on the normalized data before injecting the information on the marginals. We compare the copulas framework to IPF and to modern probabilistic approaches such as Bayesian networks, variational auto-encoders, and generative adversarial networks. We also illustrate on American Community Survey data that the method proposed allows to study the structure of the data at different geographical levels in a way that is robust to the peculiarities of the marginal distributions.

Wikipedia is one of the richest knowledge sources on the Web today. In order to facilitate navigating, searching, and maintaining its content, Wikipedia's guidelines state that all articles should be annotated with a so-called short description indicating the article's topic (e.g., the short description of beer is "Alcoholic drink made from fermented cereal grains"). Nonetheless, a large fraction of articles (ranging from 10.2% in Dutch to 99.7% in Kazakh) have no short description yet, with detrimental effects for millions of Wikipedia users. Motivated by this problem, we introduce the novel task of automatically generating short descriptions for Wikipedia articles and propose Descartes, a multilingual model for tackling it. Descartes integrates three sources of information to generate an article description in a target language: the text of the article in all its language versions, the already-existing descriptions (if any) of the article in other languages, and semantic type information obtained from a knowledge graph. We evaluate a Descartes model trained for handling 25 languages simultaneously, showing that it beats baselines (including a strong translation-based baseline) and performs on par with monolingual models tailored for specific languages. A human evaluation on three languages further shows that the quality of Descartes's descriptions is largely indistinguishable from that of human-written descriptions; e.g., 91.3% of our English descriptions (vs. 92.1% of human-written descriptions) pass the bar for inclusion in Wikipedia, suggesting that Descartes is ready for production, with the potential to support human editors in filling a major gap in today's Wikipedia across languages.

Ensuring trustworthiness in machine learning (ML) models is a multi-dimensional task. In addition to the traditional notion of predictive performance, other notions such as privacy, fairness, robustness to distribution shift, adversarial robustness, interpretability, explainability, and uncertainty quantification are important considerations to evaluate and improve (if deficient). However, these sub-disciplines or 'pillars' of trustworthiness have largely developed independently, which has limited us from understanding their interactions in real-world ML pipelines. In this paper, focusing specifically on compositions of functions arising from the different pillars, we aim to reduce this gap, develop new insights for trustworthy ML, and answer questions such as the following. Does the composition of multiple fairness interventions result in a fairer model compared to a single intervention? How do bias mitigation algorithms for fairness affect local post-hoc explanations? Does a defense algorithm for untargeted adversarial attacks continue to be effective when composed with a privacy transformation? Toward this end, we report initial empirical results and new insights from 9 different compositions of functions (or pipelines) on 7 real-world datasets along two trustworthy dimensions - fairness and explainability. We also report progress, and implementation choices, on an extensible composer tool to encourage the combination of functionalities from multiple pillars. To-date, the tool supports bias mitigation algorithms for fairness and post-hoc explainability methods. We hope this line of work encourages the thoughtful consideration of multiple pillars when attempting to formulate and resolve a trustworthiness problem.

An object-oriented approach to implementing artificial neural networks is introduced in this article. The networks obtained in this way are highly connected in that they admit edges between nodes in any layers of the network, and dynamic, in that the insertion, or deletion, of nodes, edges or layers of nodes can be effected in a straightforward way. In addition, the activation functions of nodes need not be uniform within layers, and can also be changed within individual nodes. Methods for implementing the feedforward step and the backpropagation technique in such networks are presented here. Methods for creating networks, for implementing the various dynamic properties and for saving and recreating networks are also described.