Recent works demonstrated the existence of a double-descent phenomenon for the generalization error of neural networks, where highly overparameterized models escape overfitting and achieve good test performance, at odds with the standard bias-variance trade-off described by statistical learning theory. In the present work, we explore a link between this phenomenon and the increase of complexity and sensitivity of the function represented by neural networks. In particular, we study the Boolean mean dimension (BMD), a metric developed in the context of Boolean function analysis. Focusing on a simple teacher-student setting for the random feature model, we derive a theoretical analysis based on the replica method that yields an interpretable expression for the BMD, in the high dimensional regime where the number of data points, the number of features, and the input size grow to infinity. We find that, as the degree of overparameterization of the network is increased, the BMD reaches an evident peak at the interpolation threshold, in correspondence with the generalization error peak, and then slowly approaches a low asymptotic value. The same phenomenology is then traced in numerical experiments with different model classes and training setups. Moreover, we find empirically that adversarially initialized models tend to show higher BMD values, and that models that are more robust to adversarial attacks exhibit a lower BMD.

With the rapid proliferation of textual data, predicting long texts has emerged as a significant challenge in the domain of natural language processing. Traditional text prediction methods encounter substantial difficulties when grappling with long texts, primarily due to the presence of redundant and irrelevant information, which impedes the model's capacity to capture pivotal insights from the text. To address this issue, we introduce a novel approach to long-text classification and prediction. Initially, we employ embedding techniques to condense the long texts, aiming to diminish the redundancy therein. Subsequently,the Bidirectional Encoder Representations from Transformers (BERT) embedding method is utilized for text classification training. Experimental outcomes indicate that our method realizes considerable performance enhancements in classifying long texts of Preferential Trade Agreements. Furthermore, the condensation of text through embedding methods not only augments prediction accuracy but also substantially reduces computational complexity. Overall, this paper presents a strategy for long-text prediction, offering a valuable reference for researchers and engineers in the natural language processing sphere.

Compliance with the European Union's Platform-to-Business (P2B) Regulation is challenging for online platforms, and assessing their compliance can be difficult for public authorities. This is partly due to the lack of automated tools for assessing the information (e.g., software documentation) platforms provide concerning ranking transparency. Our study tackles this issue in two ways. First, we empirically evaluate the compliance of six major platforms (Amazon, Bing, Booking, Google, Tripadvisor, and Yahoo), revealing substantial differences in their documentation. Second, we introduce and test automated compliance assessment tools based on ChatGPT and information retrieval technology. These tools are evaluated against human judgments, showing promising results as reliable proxies for compliance assessments. Our findings could help enhance regulatory compliance and align with the United Nations Sustainable Development Goal 10.3, which seeks to reduce inequality, including business disparities, on these platforms.

In this note we use the State of the Union Address (SOTU) dataset from Kaggle to make some surprising (and some not so surprising) observations pertaining to the general timeline of American history, and the character and nature of the addresses themselves. Our main approach is using vector embeddings, such as BERT (DistilBERT) and GPT-2. While it is widely believed that BERT (and its variations) is most suitable for NLP classification tasks, we find out that GPT-2 in conjunction with nonlinear dimension reduction methods such as UMAP provide better separation and stronger clustering. This makes GPT-2 + UMAP an interesting alternative. In our case, no model fine-tuning is required, and the pre-trained out-of-the-box GPT-2 model is enough. We also used a fine-tuned DistilBERT model for classification detecting which President delivered which address, with very good results (accuracy 93% - 95% depending on the run). An analogous task was performed to determine the year of writing, and we were able to pin it down to about 4 years (which is a single presidential term). It is worth noting that SOTU addresses provide relatively small writing samples (with about 8'000 words on average, and varying widely from under 2'000 words to more than 20'000), and that the number of authors is relatively large (we used SOTU addresses of 42 US presidents). This shows that the techniques employed turn out to be rather efficient, while all the computations described in this note can be performed using a single GPU instance of Google Colab. The accompanying code is available on GitHub.

The analysis of (social) networks and multi-agent systems is a central theme in Artificial Intelligence. Some line of research deals with finding groups of agents that could work together to achieve a certain goal. To this end, different notions of so-called clusters or communities have been introduced in the literature of graphs and networks. Among these, defensive alliance is a kind of quantitative group structure. However, all studies on the alliance so for have ignored one aspect that is central to the formation of alliances on a very intuitive level, assuming that the agents are preconditioned concerning their attitude towards other agents: they prefer to be in some group (alliance) together with the agents they like, so that they are happy to help each other towards their common aim, possibly then working against the agents outside of their group that they dislike. Signed networks were introduced in the psychology literature to model liking and disliking between agents, generalizing graphs in a natural way. Hence, we propose the novel notion of a defensive alliance in the context of signed networks. We then investigate several natural algorithmic questions related to this notion. These, and also combinatorial findings, connect our notion to that of correlation clustering, which is a well-established idea of finding groups of agents within a signed network. Also, we introduce a new structural parameter for signed graphs, signed neighborhood diversity snd, and exhibit a parameterized algorithm that finds a smallest defensive alliance in a signed graph.

United States Secretary of State Antony Blinken on Monday said the US is committed to deeper relations with Africa despite global crises as he opened a four-country tour of the continent. Blinken is touring four democracies on the Atlantic Coast – Cape Verde, Ivory Coast, Nigeria and Angola – as security deteriorates in the Sahel and doubts grow about a key US base in neighbouring coup-hit Niger. US President Joe Biden welcomed leaders from Africa in 2022 in a show of newfound attention to the continent. But he did not visit Africa last year as promised. Blinken nonetheless quoted Biden as he vowed, "We are all in when it comes to Africa."

OpenAI, the maker of ChatGPT, has quietly changed its rules and removed a ban on using the chatbot and its other AI tools for military purposes - and revealed that it is already working with the Department of Defense. Experts have previously voiced fears that AI could escalate conflicts around the world thanks to'slaughterbots' which can kill without any human intervention. The rule change, which occurred after Wednesday last week, removed a sentence which said that the company would not permit usage of models for'activity that has high risk of physical harm, including: weapons development, military and warfare.' The spokesman said: 'Our policy does not allow our tools to be used to harm people, develop weapons, for communications surveillance, or to injure others or destroy property. 'There are, however, national security use cases that align with our mission.

In a very relatable moment, NASA struggled for three months to get the lid off its asteroid sample container, having sent it into deep(ish) space and back. Same, NASA, same: I've struggled with jars of pickles. The space agency was finally able to get into the asteroid Bennu sample container last week and published a high-resolution image of its Touch-and-Go-Sample Acquisition Mechanism (TAGSAM) on Friday, revealing a delightful array of dust and rocks, scraped off Bennu by spacecraft OSIRIS-REx. The TAGSAM lives in a special glove compartment to prevent the sample from being contaminated, and only certain tools are approved for use with it. The team eventually had to develop new tools to open the fasteners. Tapping it on the side of the kitchen counter did not work.

While tech has played a major role in campaigns and political discourse over the past 15 years or so, and candidates and political parties have long tried to make use of big data to learn about and target voters, the past offers limited insight into where we are now. The ground is shifting incredibly quickly at technology's intersection with business, information, and media. So this week I want to run down three of the most important technology trends in the election space that you should stay on top of. Perhaps unsurprisingly, generative AI takes the top spot on our list. Without a doubt, AI that generates text or images will turbocharge political misinformation.

Many years before ChatGPT was released, my research group, the University of Cambridge Social Decision-Making Laboratory, wondered whether it was possible to have neural networks generate misinformation. To achieve this, we trained ChatGPT's predecessor, GPT-2, on examples of popular conspiracy theories and then asked it to generate fake news for us. It gave us thousands of misleading but plausible-sounding news stories. A few examples: "Certain Vaccines Are Loaded With Dangerous Chemicals and Toxins," and "Government Officials Have Manipulated Stock Prices to Hide Scandals." The question was, would anyone believe these claims?