Hasil untuk "Auxiliary sciences of history"

Gustavo Barrientos, Gustavo Barrientos, Luciana Catella et al.

The aim of this paper is to present and discuss an approach to address the spatial variation in the degree and type of omnivory exhibited by human populations that inhabited the temperate zone of South America east of the Andes (30°-56° S) during the late Holocene. This approach is based on the interpolation mapping of transformed isotopic niches, understood as the position occupied by an individual or group of individuals in a space that results from transforming one or more of the delta (δ) variables that specify the original isotopic niche (e.g., δ15N [‰]) into derived variables such as trophic position (TP). Our results indicate a strong spatial structuring of both transformed isotopic niches and three omnivory categories (OC I, OC II, and OC III), defined by ranges of TP values (i.e., 2.0–2.99; 3.0–3.99; ≥4.0). Among the factors that likely structured spatial variation in the degree and type of omnivory are those characterizing the physical environment (e.g., net primary productivity or NPP, effective temperature or ET) and the biotic environment (e.g., differential distribution of marine biota). Since these factors have confounding effects, it is difficult to distinguish, given our current state of knowledge, which is the most important. For this reason, we conclude that macroecological analyses are needed that go beyond pattern recognition to address the identification and explanation of underlying processes.

Jong Hyeon Lee, Jongmin Kim, Heesang Lee et al.

A large class of spatial models contains intractable normalizing functions, such as spatial lattice models, interaction spatial point processes, and social network models. Bayesian inference for such models is challenging since the resulting posterior distribution is doubly intractable. Although auxiliary variable MCMC (AVM) algorithms are known to be the most practical, they are computationally expensive due to the repeated auxiliary variable simulations. To address this, we propose delayed-acceptance AVM (DA-AVM) methods, which can reduce the number of auxiliary variable simulations. The first stage of the kernel uses a cheap surrogate to decide whether to accept or reject the proposed parameter value. The second stage guarantees detailed balance with respect to the posterior. The auxiliary variable simulation is performed only on the parameters accepted in the first stage. We construct various surrogates specifically tailored for doubly intractable problems, including subsampling strategy, Gaussian process emulation, and frequentist estimator-based approximation. We validate our method through simulated and real data applications, demonstrating its practicality for complex spatial models.

Alessio Quercia, Erenus Yildiz, Zhuo Cao et al.

Monocular depth estimation (MDE) is a challenging task in computer vision, often hindered by the cost and scarcity of high-quality labeled datasets. We tackle this challenge using auxiliary datasets from related vision tasks for an alternating training scheme with a shared decoder built on top of a pre-trained vision foundation model, while giving a higher weight to MDE. Through extensive experiments we demonstrate the benefits of incorporating various in-domain auxiliary datasets and tasks to improve MDE quality on average by ~11%. Our experimental analysis shows that auxiliary tasks have different impacts, confirming the importance of task selection, highlighting that quality gains are not achieved by merely adding data. Remarkably, our study reveals that using semantic segmentation datasets as Multi-Label Dense Classification (MLDC) often results in additional quality gains. Lastly, our method significantly improves the data efficiency for the considered MDE datasets, enhancing their quality while reducing their size by at least 80%. This paves the way for using auxiliary data from related tasks to improve MDE quality despite limited availability of high-quality labeled data. Code is available at https://jugit.fz-juelich.de/ias-8/mdeaux.

Yahya Badran, Christine Preisach

Personalized recommendation is a key feature of intelligent tutoring systems, typically relying on accurate models of student knowledge. Knowledge Tracing (KT) models enable this by estimating a student's mastery based on their historical interactions. Many KT models rely on human-annotated knowledge concepts (KCs), which tag each exercise with one or more skills or concepts believed to be necessary for solving it. However, these KCs can be incomplete, error-prone, or overly general. In this paper, we propose a deep learning model that learns sparse binary representations of exercises, where each bit indicates the presence or absence of a latent concept. We refer to these representations as auxiliary KCs. These representations capture conceptual structure beyond human-defined annotations and are compatible with both classical models (e.g., BKT) and modern deep learning KT architectures. We demonstrate that incorporating auxiliary KCs improves both student modeling and adaptive exercise recommendation. For student modeling, we show that augmenting classical models like BKT with auxiliary KCs leads to improved predictive performance. For recommendation, we show that using auxiliary KCs enhances both reinforcement learning-based policies and a simple planning-based method (expectimax), resulting in measurable gains in student learning outcomes within a simulated student environment.

Jingxuan Wei, Caijun Jia, Qi Chen et al.

Mathematical geometric reasoning is essential for scientific discovery and educational development, requiring precise logic and rigorous formal verification. While recent advances in Multimodal Large Language Models (MLLMs) have improved reasoning tasks, existing models typically struggle with formal geometric reasoning, particularly when dynamically constructing and verifying auxiliary geometric elements. To address these challenges, we introduce Geoint-R1, a multimodal reasoning framework designed to generate formally verifiable geometric solutions from textual descriptions and visual diagrams. Geoint-R1 uniquely integrates auxiliary elements construction, formal reasoning represented via Lean4, and interactive visualization. To systematically evaluate and advance formal geometric reasoning, we propose the Geoint benchmark, comprising 1,885 rigorously annotated geometry problems across diverse topics such as plane, spatial, and solid geometry. Each problem includes structured textual annotations, precise Lean4 code for auxiliary constructions, and detailed solution steps verified by experts. Extensive experiments demonstrate that Geoint-R1 significantly surpasses existing multimodal and math-specific reasoning models, particularly on challenging problems requiring explicit auxiliary element constructions.

Anna Rygula, Anna Klisińska-Kopacz, Paulina Krupska-Wolas et al.

Amber has been used to create decorative items for centuries, but its degradation presents challenges for conservators. This study identifies substances historically used to protect amber objects, especially those from 17th and 18th century Gdansk workshops. Despite their historical value, information on amber conservation is scarce. Traditional substances are noted, but their exact compositions and effects on amber remain unclear. Synthetic resins, introduced in the late 19th century, also degrade, complicating conservation due to their removal difficulty and interference with amber identification. This research aimed to develop methods for detecting and analyzing protective coatings on amber objects using macroscopic and microscopic techniques. Initial methods included analytical photography under visible and UV light and reflectance imaging spectroscopy (RIS) to assess the surface. Raman spectroscopy (RS) and X-ray fluorescence spectroscopy (XRF) were used for detailed analysis. RS provided precise layer-specific information but was sensitive to surface conditions, while XRF quickly identified inorganic compounds but not organic materials. Examining amber objects from Polish collections using this methodology revealed various protective substances, including synthetic resins and nitrocellulose varnishes. This research contributes to amber conservation by proposing a comprehensive material analysis approach, essential for developing effective conservation strategies for these historic objects.

Seonghyeon Lee, Suyeon Kim, Joonwon Jang et al.

We study the code generation behavior of instruction-tuned models built on top of code pre-trained language models when they could access an auxiliary function to implement a function. We design several ways to provide auxiliary functions to the models by adding them to the query or providing a response prefix to incorporate the ability to utilize auxiliary functions with the instruction-following capability. Our experimental results show the effectiveness of combining the base models' auxiliary function utilization ability with the instruction following ability. In particular, the performance of adopting our approaches with the open-sourced language models surpasses that of the recent powerful proprietary language models, i.e., gpt-4o.

María E. Pedreschi M., Adys Pereira de Herrera

Los bancos son instituciones vitales para el funcionamiento de la economía moderna, permitiendo a empresas y personas el acceso a recursos y servicios financieros para la consecución de sus fines de inversión y consumo, propiciando el desarrollo macroeconómico y la disminución de la pobreza de los países. La bancarización se ha convertido en la principal herramienta de inclusión financiera, promovida por los estados y organismos internacionales. En el 2018 y 2019 se realizó una investigación descriptiva-transversal con el objetivo de medir algunas variables relacionadas con la bancarización de la población universitaria en el distrito de Chitré, desde el enfoque de demanda y con ello, describir sus principales características. La muestra fue estratificada contemplando tres universidades de mayor matrícula. Se observó que los niveles de bancarización se diferencian entre los estratos, pues en los profesores y administrativos estos son mayores (97.1%) que el de los estudiantes (60.0%). Los universitarios se bancarizan tanto a través del ahorro como del crédito (81% y 16%, respectivamente), poseen más de una cuenta de ahorro, en más de un banco, sobre todo para realizar transacciones de pagos. Los profesores y administrativos manejan en promedio 2.3 créditos y los estudiantes, 1.4. En la actualidad, se espera que los indicadores de bancarización sean mayores, tanto por la dinámica de avance de la tecnología financiera, como por los impactos positivos de la pandemia de Covid-19 mostrados por el aumento de la tenencia de cuentas de ahorro y su uso electrónico para realizar transacciones. Se concluye que la población universitaria puede ser vista como un nicho de mercado importante, sustentando el diseño de productos y servicios y programas de educación financiera focalizados a este segmento de la población general.

Luis Henrique de Camargo

Este artigo objetiva verificar a relação da sociedade com a natureza, e os seus fluxos energéticos termodinâmicos, como elemento evolutivo, gerando totalização e sendo analisados pela(s) flecha(s) do espaço-tempo.  Este artigo verificará também, como este processo influencia na formação do Antropoceno. Neste sentido, serão aplicados os princípios nascidos após o advento da mecânica quântica, à análise espaço-temporal da physis (que integra sociedade-natureza). O artigo verifica, também, como cada forma-conteúdo, de forma singular, contribui energeticamente para o desenvolvimento da sua flecha do espaço-tempo. Assim, será verificado também como o processo produtivo atual associa-se, em geral, ao rompimento dos estados de homeostase, no balanço energético do processo de troca entre energia e matéria, e como o antigo padrão de relativa estabilidade, que caracterizava o Holoceno, vem sendo substituído pela desordem, que está na base do surgimento do Antropoceno.

Piotr Kania-Kaniowski

(Defensor civitatis. Legal protection authority in the later Roman Empire): In the 4th century, an official with the title defensor civitatis became widespread in the Roman Empire. His task was to protect the poor from administrative injustices. Noteworthy were his powers regarding taxation, which served the state’s fiscal goals by protecting sources of tax revenue. Initially, the defensor enjoyed a certain level of independence, but over time the position became subordinated to the local elites. In the 5th and 6th centuries, the original function of this official was distorted.

Tomoyuki Morimae, Barak Nehoran, Takashi Yamakawa

We show the following unconditional results on quantum commitments in two related yet different models: 1. We revisit the notion of quantum auxiliary-input commitments introduced by Chailloux, Kerenidis, and Rosgen (Comput. Complex. 2016) where both the committer and receiver take the same quantum state, which is determined by the security parameter, as quantum auxiliary inputs. We show that computationally-hiding and statistically-binding quantum auxiliary-input commitments exist unconditionally, i.e., without relying on any unproven assumption, while Chailloux et al. assumed a complexity-theoretic assumption, ${\bf QIP}\not\subseteq{\bf QMA}$. On the other hand, we observe that achieving both statistical hiding and statistical binding at the same time is impossible even in the quantum auxiliary-input setting. To the best of our knowledge, this is the first example of unconditionally proving computational security of any form of (classical or quantum) commitments for which statistical security is impossible. As intermediate steps toward our construction, we introduce and unconditionally construct post-quantum sparse pseudorandom distributions and quantum auxiliary-input EFI pairs which may be of independent interest. 2. We introduce a new model which we call the common reference quantum state (CRQS) model where both the committer and receiver take the same quantum state that is randomly sampled by an efficient setup algorithm. We unconditionally prove that there exist statistically hiding and statistically binding commitments in the CRQS model, circumventing the impossibility in the plain model. We also discuss their applications to zero-knowledge proofs, oblivious transfers, and multi-party computations.

Nurit Oliszewski, Jorge G. Martínez, Guillermo Arreguez et al.

Cueva de Los Corrales 1 se ubica en la quebrada de Los Corrales (El Infiernillo, Tucumán). Es un sitio arqueológico complejo tanto en su espacialidad como en su temporalidad ya que presenta tres sectores de uso ‒dos en el interior (cueva propiamente dicha y morteros fijos) y uno en el exterior (alero)‒ en varios momentos de ocupación entre ca. 3000 y 600 años AP. Se presenta una síntesis de la información generada que incluye el examen de distintas materialidades, dataciones radiocarbónicas y análisis desde diferentes líneas de investigación, los cuales, en conjunto, han permitido establecer al menos cuatro eventos ocupacionales. Se evalúan los posibles usos que tuvo CC1 en los distintos momentos en que estuvo habitado y su rol en el devenir de las ocupaciones humanas de la quebrada de Los Corrales y de la región del sur de cumbres Calchaquíes-norte del sistema del Aconquija en tiempos prehispánicos.

Davide Trentacoste

Diplomatic relations between two states are never simply bilateral relations as there is always a broader context in which they can be framed and, very often, they do not arise from nothing but are the result of specific events. Like all human vicissitudes, they have a beginning, a development and an end. The relations between Medici Tuscany and Safavid Persia have unfortunately never been the subject of extensive enough studies that go beyond the few things already known from decades ago. However, the continuous search for sources and their analysis in the light of a different historiographic approach can provide a new understanding of certain events and a more precise chronological and historical framework.

Loc Vu-Quoc, Alexander Humer

Three recent breakthroughs due to AI in arts and science serve as motivation: An award winning digital image, protein folding, fast matrix multiplication. Many recent developments in artificial neural networks, particularly deep learning (DL), applied and relevant to computational mechanics (solid, fluids, finite-element technology) are reviewed in detail. Both hybrid and pure machine learning (ML) methods are discussed. Hybrid methods combine traditional PDE discretizations with ML methods either (1) to help model complex nonlinear constitutive relations, (2) to nonlinearly reduce the model order for efficient simulation (turbulence), or (3) to accelerate the simulation by predicting certain components in the traditional integration methods. Here, methods (1) and (2) relied on Long-Short-Term Memory (LSTM) architecture, with method (3) relying on convolutional neural networks. Pure ML methods to solve (nonlinear) PDEs are represented by Physics-Informed Neural network (PINN) methods, which could be combined with attention mechanism to address discontinuous solutions. Both LSTM and attention architectures, together with modern and generalized classic optimizers to include stochasticity for DL networks, are extensively reviewed. Kernel machines, including Gaussian processes, are provided to sufficient depth for more advanced works such as shallow networks with infinite width. Not only addressing experts, readers are assumed familiar with computational mechanics, but not with DL, whose concepts and applications are built up from the basics, aiming at bringing first-time learners quickly to the forefront of research. History and limitations of AI are recounted and discussed, with particular attention at pointing out misstatements or misconceptions of the classics, even in well-known references. Positioning and pointing control of a large-deformable beam is given as an example.

Yun He, Xue Feng, Cheng Cheng et al.

In many personalized recommendation scenarios, the generalization ability of a target task can be improved via learning with additional auxiliary tasks alongside this target task on a multi-task network. However, this method often suffers from a serious optimization imbalance problem. On the one hand, one or more auxiliary tasks might have a larger influence than the target task and even dominate the network weights, resulting in worse recommendation accuracy for the target task. On the other hand, the influence of one or more auxiliary tasks might be too weak to assist the target task. More challenging is that this imbalance dynamically changes throughout the training process and varies across the parts of the same network. We propose a new method: MetaBalance to balance auxiliary losses via directly manipulating their gradients w.r.t the shared parameters in the multi-task network. Specifically, in each training iteration and adaptively for each part of the network, the gradient of an auxiliary loss is carefully reduced or enlarged to have a closer magnitude to the gradient of the target loss, preventing auxiliary tasks from being so strong that dominate the target task or too weak to help the target task. Moreover, the proximity between the gradient magnitudes can be flexibly adjusted to adapt MetaBalance to different scenarios. The experiments show that our proposed method achieves a significant improvement of 8.34% in terms of NDCG@10 upon the strongest baseline on two real-world datasets. The code of our approach can be found at here: https://github.com/facebookresearch/MetaBalance

Shayan Jawed, Lars Schmidt-Thieme

We propose a novel multi-task method for quantile forecasting with shared Linear layers. Our method is based on the Implicit quantile learning approach, where samples from the Uniform distribution $\mathcal{U}(0, 1)$ are reparameterized to quantile values of the target distribution. We combine the implicit quantile and input time series representations to directly forecast multiple quantile estimations for multiple horizons jointly. Prior works have adopted a Linear layer for the direct estimation of all forecasting horizons in a multi-task learning setup. We show that following similar intuition from multi-task learning to exploit correlations among forecast horizons, we can model multiple quantile estimates as auxiliary tasks for each of the forecast horizon to improve forecast accuracy across the quantile estimates compared to modeling only a single quantile estimate. We show learning auxiliary quantile tasks leads to state-of-the-art performance on deterministic forecasting benchmarks concerning the main-task of forecasting the 50$^{th}$ percentile estimate.

Miloslav Znojil

The current applications of non-Hermitian but ${\cal PT}-$symmetric Hamiltonians $H$ cover several, mutually not too closely connected subdomains of quantum physics. Mathematically, the split between the open and closed systems can be characterized by the respective triviality and non-triviality of an auxiliary inner-product metric $Θ=Θ(H)$. With our attention restricted to the latter, mathematically more interesting unitary-evolution case we show that the intuitive but technically decisive simplification of the theory achieved via an "additional" ${\cal PCT}-$symmetry constraint upon $H$ can be given a deeper mathematical meaning via introduction of a certain second auxiliary inner product.

Bernadette Lessel, Thomas Schick Max Planck Institute for the History of Science, Berlin Mathematisches Institut et al.

A notion of differentiability is being proposed for maps between Wasserstein spaces of order 2 of smooth, connected and complete Riemannian manifolds. Due to the nature of the tangent space construction on Wasserstein spaces, we only give a global definition of differentiability, i.e. without a prior notion of pointwise differentiability. With our definition, however, we recover the expected properties of a differential. Special focus is being put on differentiability properties of pushforward maps induced by smooth maps between the underlying manifolds, and on convex mixing of differentiable maps, with an explicit construction of the differential.

Anne Marie Pessis

A Revista Clio Arqueológica, publicada interruptamente há 35 anos, incorporou várias temáticas ao seu conteúdo. Além de pré-história e arqueologia histórica, hoje se publicam trabalhos sobre restauração e preservação patrimonial e sobre arqueometria. Entretanto, para pesquisadores de outras áreas, como a física, a química, a biologia, por exemplo, uma vez que fornecem subsídios às pesquisas arqueológicas através de técnicas de análises inovadoras, criou-se na Revista o espaço Academicum Opus, no qual se entrevistam pesquisadores que trabalham em conjunto com os arqueólogos. Essa postura, que tem possibilitado novas abordagens e aportado o benefício da prova e dos dados científicos, tão necessários às ciências humanas, denotada a missão interdisciplinar da Clio Arqueológica e do seu conteúdo científico. Neste número entrevistamos o Professor Renato Athias, Antropólogo do Departamento de Antropologia da UFPE que discorreu sobre as relações entre arqueologia e antropologia. A Clio é uma publicação na qual pesquisadores de distintas áreas do conhecimento divulgam textos inéditos e correlatos a temas arqueológicos com um único objetivo: entender nosso passado a partir de vestígios materiais.

Sayed Awad Mohamed, Ruey-Lin Chang, Cassandre Hartenstein et al.