We introduce the notion of a naive global 2-ring: a functor from the opposite of the $\infty$-category of global spaces to presentably symmetric monoidal stable $\infty$-categories. By passing to global sections, every naive global 2-ring decategorifies to a multiplicative cohomology theory on global spaces, i.e. a naive global ring. We suggest when a naive global 2-ring deserves to be called \emph{genuine}. As evidence, we associate to such a global 2-ring a family of equivariant cohomology theories which satisfy a version of the change of group axioms introduced by Ginzburg, Kapranov and Vasserot. We further show that the decategorified multiplicative global cohomology theory associated to a genuine global $2$-ring canonically refines to an $\mathbb{E}_\infty$-ring object in global spectra. As we show, two interesting examples of genuine global 2-rings are given by quasi-coherent sheaves on the torsion points of an oriented spectral elliptic curve and Lurie's theory of tempered local systems. In particular, we obtain global spectra representing equivariant elliptic cohomology and tempered cohomology.
This chapter examines the global governance of artificial intelligence (AI) through the lens of the Global AI Divide, focusing on disparities in AI development, innovation, and regulation. It highlights systemic inequities in education, digital infrastructure, and access to decision-making processes, perpetuating a dependency and exclusion cycle for Global Majority countries. The analysis also explores the dominance of Western nations and corporations in shaping AI governance frameworks, which often sideline the unique priorities and contexts of the Global Majority. Additionally, this chapter identifies emerging countertrends, such as national and regional AI strategies, as potential avenues for fostering equity and inclusivity in global AI governance. The chapter concludes with actionable recommendations to democratize AI governance for Majority World countries, emphasizing the importance of systemic reforms, resource redistribution, and meaningful participation. It calls for collaborative action to ensure AI governance becomes a catalyst for shared prosperity, addressing global disparities rather than deepening them.
Biomimetic nanotechnology is a prominent research area at the meeting place of life sciences with engineering and physics: it is a continuously growing field that deals with knowledge transfer from biology to nanotechnology. Biomimetic nanotechnology is a field that has the potential to substantially support successful mastering of major global challenges. The Millennium Project was commissioned by the United Nations Secretary-General in 2002 to develop a concrete action plan for the world to reverse the grinding poverty, hunger and disease affecting billions of people. It states 15 Global Challenges: sustainable development, water, population and resources, democratization, long-term perspectives, information technology, the rich-poor gap, health, capacity to decide, peace and conflict, status of women, transnational crime, energy, science and technology and global ethics. The possible contributions to master these challenges with the help of biomimetic nanotechnology will be discussed in detail.
Global cooperation often falters despite shared objectives, as misaligned interests and unequal incentives undermine collective efforts, such as those in international climate change collaborations. To tackle this issue, this paper introduces a multi-level game-theoretic model to analyze the dynamics of complex interactions within hierarchical systems. The model consists of global, local, and pairwise games, and two strategy types, binary and level-based strategies, are explored under varying parameter conditions. Using computational simulations and numerical analysis, we examine how factors across different levels influence player decisions, game dynamics and population phase transitions during the evolutionary process. Our findings reveal that although the increase of profit rates at local and pairwise games enhances cooperation within the population, the global game exerts minimal influence on player decisions and population states under both strategy settings. Particularly, analytical and simulation results show that, under binary strategies, global profit does not influence localized decision-making of players, while under level-based strategies, players cooperating at the glo
Chest X-ray interpretation is one of the most frequently performed diagnostic tasks in medicine and a primary target for AI development, yet current vision-language models are primarily trained on datasets of paired images and reports, not the cognitive processes and visual attention that underlie clinical reasoning. Here, we present CheXthought, a global, multimodal resource containing 103,592 chain-of-thought reasoning traces and 6,609,082 synchronized visual attention annotations across 50,312 multi-read chest X-rays from 501 radiologists in 71 countries. Our analysis reveals clinical reasoning patterns in how experts deploy distinct visual search strategies, integrate clinical context, and communicate uncertainty. We demonstrate the clinical utility of CheXthought across four dimensions. First, CheXthought reasoning significantly outperforms state-of-the-art vision-language model chain-of-thought in factual accuracy and spatial grounding. Second, visual attention data used as an inference-time hint recovers missed findings and significantly reduces hallucinations. Third, vision-language models trained on CheXthought data achieve significantly stronger pathology classification,
Federated learning (FL) with a single global server framework is currently a popular approach for training machine learning models on decentralized environment, such as mobile devices and edge devices. However, the centralized server architecture poses a risk as any challenge on the central/global server would result in the failure of the entire system. To minimize this risk, we propose a novel federated learning framework that leverages the deployment of multiple global servers. We posit that implementing multiple global servers in federated learning can enhance efficiency by capitalizing on local collaborations and aggregating knowledge, and the error tolerance in regard to communication failure in the single server framework would be handled. We therefore propose a novel framework that leverages the deployment of multiple global servers. We conducted a series of experiments using a dataset containing the event history of electric vehicle (EV) charging at numerous stations. We deployed a federated learning setup with multiple global servers and client servers, where each client-server strategically represented a different region and a global server was responsible for aggregating
Text generation, a key component in applications such as dialogue systems, relies on decoding algorithms that sample strings from a language model distribution. Traditional methods, such as top-$k$ and top-$π$, apply local normalisation to the model's output distribution, which can distort it. In this paper, we investigate the effect of this distortion by introducing globally-normalised versions of these decoding methods. Additionally, we propose an independent Metropolis-Hastings algorithm to approximate sampling from globally-normalised distributions without explicitly computing them. Our empirical analysis compares the performance of local and global normalisation across two decoding algorithms (top-$k$ and top-$π$) with various hyperparameters, using Pythia language models. Results show that, in most configurations, global decoding performs worse than the local decoding version of the same algorithms -- despite preserving the distribution's integrity. Our results suggest that distortion is an important feature of local decoding algorithms.
This paper examines how international AI governance frameworks address gender issues and gender-based harms. The analysis covers binding regulations, such as the EU AI Act; soft law instruments, like the UNESCO Recommendations on AI Ethics; and global initiatives, such as the Global Partnership on AI (GPAI). These instruments reveal emerging trends, including the integration of gender concerns into broader human rights frameworks, a shift toward explicit gender-related provisions, and a growing emphasis on inclusivity and diversity. Yet, some critical gaps persist, including inconsistent treatment of gender across governance documents, limited engagement with intersectionality, and a lack of robust enforcement mechanisms. However, this paper argues that effective AI governance must be intersectional, enforceable, and inclusive. This is key to moving beyond tokenism toward meaningful equity and preventing reinforcement of existing inequalities. The study contributes to ethical AI debates by highlighting the importance of gender-sensitive governance in building a just technological future.
In the face of socioeconomic challenges, this paper develops and empirically demonstrates the Gondauri Index (GI) as a reproducible diagnostics-first composite framework for benchmarking macro-financial resilience across heterogeneous economies on a unified 0-100 scale. The GI addresses a key limitation of conventional surveillance dashboards: resilience is multi-dimensional and only partially substitutable, so strength in one area cannot sustainably offset fragility in another. The index integrates three interpretable pillars: Inequality Resilience Score (IRS), Liquidity and Systemic Resilience (LNSR), and Inflation Forecast Coherence (IFC). Cross-country comparability is ensured through robust percentile normalization (p5-p95), a consistent annual country-year design, and explicit missing-data handling via component-level weight renormalization. Empirically, the paper provides a 2024 benchmark snapshot and dynamic evidence for 2005-2024 using 5-year rolling diagnostics and Delta log(GI) contribution decomposition, allowing transparent attribution of resilience changes to pillar-level drivers. A forward-looking extension constructs 2026-2030 scenario pathways and introduces a bind
Modern vision models have achieved remarkable success in benchmarks where local features provide critical information about the target. There is now a growing interest in tackling tasks requiring more global reasoning, where local features do not provide significant information. Minsky and Papert put forward such tasks in 1969 with their connectivity study, exposing the limitations of the perceptron model. In this paper, we introduce an expanded set of global visual datasets involving graphs, strings, mazes, and image grids. We show that large vision models still struggle to learn these tasks efficiently. Similarly, state-of-the-art multi-modal LLMs perform poorly on these datasets. We explain this learning inefficiency by means of the 'globality degree' measure. To mitigate this, we propose a method called chain-of-sketch (CoS). Similar to the chain-of-thought and scratchpad techniques used in language models, CoS breaks the original task into intermediate visual steps to help learn a complex task. In addition, we show that not all CoS strategies perform equally well. Our key insight is to impose a Markovian structure on the CoS frames. This leads to the introduction of 'inductive
The well-posedness of three dimensional Prandtl equation is an outstanding open problem despite of the study in analytic and Gevrey function spaces. This problem is raised as the third open problem in the classical monograph by Oleinik and Samokhin [43]. The paper aims to address this open problem in the steady case by introducing a novel approach to establish the global stability of background profile that includes the celebrated Blasius solutions, which is of particular interest in light of the recent affirmation of Prandtl's ansatz in two dimensional steady setting by Iyer and Masmoudi [32]. In three dimensions, the well-established analytic approaches for two dimensional setting can not be applied because of the appearance of the secondary flow. Rather than employing cancellation mechanisms or coordinate transforms, we introduce new intrinsic vector fields featuring curvature-type commutators and establish vector field-based maximum principles through pointwise and integral estimates to address the loss of tangential derivatives.
In this paper we develop the definition of a global orthogonal spectrum and its unitary version. It relates $G-$equivariant spectra by equivariant weak equivalence in a coherent way. This category of global spectra has a model structure Quillen equivalent to the global model structure on orthogonal spectra. We also show that there is a large family of equivariant cohomology theories, including quasi-elliptic cohomology, that can be globalized in the new context. Starting from one global ring spectrum, we can construct infinitely many distinct global ring spectra. Moreover, in light of the results in this paper, we ask whether we have the conjecture that the globalness of a cohomology theory is completely determined by the formal component of its divisible group and when the $\acute{e}$tale component of it varies the globalness does not change.
Modern economic systems face unprecedented socioeconomic challenges, making systemic resilience and effective liquidity flow management essential. Traditional models such as CAPM, VaR, and GARCH often fail to reflect real market fluctuations and extreme events. This study develops and validates an innovative mathematical model based on the Navier-Stokes equations, aimed at the quantitative assessment, forecasting, and simulation of liquidity flows and systemic risks. The model incorporates 13 macroeconomic and financial parameters, including liquidity velocity, market pressure, internal stress, stochastic fluctuations, and risk premiums, all based on real data and formally included in the modified equation. The methodology employs econometric testing, Fourier analysis, stochastic simulation, and AI-based calibration to enable dynamic testing and forecasting. Simulation-based sensitivity analysis evaluates the impact of parameter changes on financial balance. The model is empirically tested using Georgian macroeconomic and financial data from 2010-2024, including GDP, inflation, the Gini index, CDS spreads, and LCR metrics. Results show that the model effectively describes liquidity
Humans can discern scene-independent features of objects across various environments, allowing them to swiftly identify objects amidst changing factors such as lighting, perspective, size, and position and imagine the complete images of the same object in diverse settings. Existing object-centric learning methods only extract scene-dependent object-centric representations, lacking the ability to identify the same object across scenes as humans. Moreover, some existing methods discard the individual object generation capabilities to handle complex scenes. This paper introduces a novel object-centric learning method to empower AI systems with human-like capabilities to identify objects across scenes and generate diverse scenes containing specific objects by learning a set of global object-centric representations. To learn the global object-centric representations that encapsulate globally invariant attributes of objects (i.e., the complete appearance and shape), this paper designs a Disentangled Slot Attention module to convert the scene features into scene-dependent attributes (such as scale, position and orientation) and scene-independent representations (i.e., appearance and shape
This paper studies the global well-posedness and optimal decay estimates to the Oldroyd-B model in $\mathbb R^d$ ($d\geq2$). By utilizing the special structure of this system, we give a simplified proof to the global existence of solutions for the case of initial data small in critical Besov spaces and non-small coupling parameters. Moreover, the optimal decay rates of the solutions under minimal small assumption on the initial data are established by fully making use of the effect of velocity dissipation and damping mechanism.
Global and local relational reasoning enable scene understanding models to perform human-like scene analysis and understanding. Scene understanding enables better semantic segmentation and object-to-object interaction detection. In the medical domain, a robust surgical scene understanding model allows the automation of surgical skill evaluation, real-time monitoring of surgeon's performance and post-surgical analysis. This paper introduces a globally-reasoned multi-task surgical scene understanding model capable of performing instrument segmentation and tool-tissue interaction detection. Here, we incorporate global relational reasoning in the latent interaction space and introduce multi-scale local (neighborhood) reasoning in the coordinate space to improve segmentation. Utilizing the multi-task model setup, the performance of the visual-semantic graph attention network in interaction detection is further enhanced through global reasoning. The global interaction space features from the segmentation module are introduced into the graph network, allowing it to detect interactions based on both node-to-node and global interaction reasoning. Our model reduces the computation cost compa
In this paper, we present the Global Multimedia Deepfake Detection held concurrently with the Inclusion 2024. Our Multimedia Deepfake Detection aims to detect automatic image and audio-video manipulations including but not limited to editing, synthesis, generation, Photoshop,etc. Our challenge has attracted 1500 teams from all over the world, with about 5000 valid result submission counts. We invite the top 20 teams to present their solutions to the challenge, from which the top 3 teams are awarded prizes in the grand finale. In this paper, we present the solutions from the top 3 teams of the two tracks, to boost the research work in the field of image and audio-video forgery detection. The methodologies developed through the challenge will contribute to the development of next-generation deepfake detection systems and we encourage participants to open source their methods.
Generative AI technologies are gaining unprecedented popularity, causing a mix of excitement and apprehension through their remarkable capabilities. In this paper, we study the challenges associated with deploying synthetic data, a subfield of Generative AI. Our focus centers on enterprise deployment, with an emphasis on privacy concerns caused by the vast amount of personal and highly sensitive data. We identify 40+ challenges and systematize them into five main groups -- i) generation, ii) infrastructure & architecture, iii) governance, iv) compliance & regulation, and v) adoption. Additionally, we discuss a strategic and systematic approach that enterprises can employ to effectively address the challenges and achieve their goals by establishing trust in the implemented solutions.
Background: Despite the consensus that vaccines play an important role in combating the global spread of infectious diseases, vaccine inequity is still rampant with deep-seated mentality of self-priority. This study aims to evaluate the existence and possible outcomes of a more equitable global vaccine distribution and explore a concrete incentive mechanism that promotes vaccine equity. Methods: We design a metapopulation epidemiological model that simultaneously considers global vaccine distribution and human mobility, which is then calibrated by the number of infections and real-world vaccination records during COVID-19 pandemic from March 2020 to July 2021. We explore the possibility of the enlightened self-interest incentive mechanism, i.e., improving one's own epidemic outcomes by sharing vaccines with other countries, by evaluating the number of infections and deaths under various vaccine sharing strategies using the proposed model. To understand how these strategies affect the national interests, we distinguish the imported and local cases for further cost-benefit analyses that rationalize the enlightened self-interest incentive mechanism behind vaccine sharing. ...
Many approaches for addressing Global Optimization problems typically rely on relaxations of nonlinear constraints over specific mathematical primitives. This is restricting in applications with constraints that are black-box, implicit or consist of more general primitives. Trying to address such limitations, Bertsimas and Ozturk (2023) proposed OCTHaGOn as a way of solving black-box global optimization problems by approximating the nonlinear constraints using hyperplane-based Decision-Trees and then using those trees to construct a unified mixed integer optimization (MIO) approximation of the original problem. We provide extensions to this approach, by (i) approximating the original problem using other MIO-representable ML models besides Decision Trees, such as Gradient Boosted Trees, Multi Layer Perceptrons and Suport Vector Machines, (ii) proposing adaptive sampling procedures for more accurate machine learning-based constraint approximations, (iii) utilizing robust optimization to account for the uncertainty of the sample-dependent training of the ML models, and (iv) leveraging a family of relaxations to address the infeasibilities of the final MIO approximation. We then test t