Laboratory confirmation of infection as demonstrated by: Hepatitis B surface antigen (HBsAg) positive PLUS immunoglobulin M antibody to hepatitis B core antigen (anti-HBc IgM) positive in the context of a compatible clinical history or probable exposure. OR Loss of HBsAg within a six month period after testing HBsAg positive in the context of a compatible clinical history or probable exposure AND in the absence of recent history of Hepatitis B virus (HBV) vaccination. OR Positive HBV DNA by polymerase chain reaction (PCR) followed by HBsAg and HBV core IgM antibody sero-conversion in the context of a compatible clinical history or probable exposure.
We apply, for the first time, Bayes Linear Emulation and History Matching to the calibration of non-perturbative models in Monte Carlo event generators. In contrast to the usual approach of "Monte Carlo tuning", History Matching does not result in best-fit plus ellipsoidal parameter uncertainty estimates but instead identifies all parameter space regions that are consistent with data. This approach leads to a systematic and robust quantification of parametric uncertainties in the models, especially in those challenging cases where different, possibly disjoint, regions of parameter space deliver similar results, which are usually not properly treated with current methodology. We highlight the power of this method with the hadronisation models available through Sherpa: the built-in cluster fragmentation Ahadic and string fragmentation through an interface to Pythia.
Traditional approaches to studying decision-making in neuroscience focus on simplified behavioral tasks where animals perform repetitive, stereotyped actions to receive explicit rewards. While informative, these methods constrain our understanding of decision-making to short timescale behaviors driven by explicit goals. In natural environments, animals exhibit more complex, long-term behaviors driven by intrinsic motivations that are often unobservable. Recent works in time-varying inverse reinforcement learning (IRL) aim to capture shifting motivations in long-term, freely moving behaviors. However, a crucial challenge remains: animals make decisions based on their history, not just their current state. To address this, we introduce SWIRL (SWitching IRL), a novel framework that extends traditional IRL by incorporating time-varying, history-dependent reward functions. SWIRL models long behavioral sequences as transitions between short-term decision-making processes, each governed by a unique reward function. SWIRL incorporates biologically plausible history dependency to capture how past decisions and environmental contexts shape behavior, offering a more accurate description of animal decision-making. We apply SWIRL to simulated and real-world animal behavior datasets and show that it outperforms models lacking history dependency, both quantitatively and qualitatively. This work presents the first IRL model to incorporate history-dependent policies and rewards to advance our understanding of complex, naturalistic decision-making in animals.
Ewoud Wempe, Amina Helmi, Simon D. M. White
et al.
We study the mass growth histories of the halos of Milky Way and M31 analogues formed in constrained cosmological simulations of the Local Group. These simulations constitute a fair and representative set of $Λ$CDM realisations conditioned on properties of the main Local Group galaxies, such as their masses, relative separation, dynamics and environment. Comparing with isolated analogues extracted from the TNG dark-matter-only simulations, we find that while our M31 halos have a comparable mass growth history to their isolated counterparts, our Milky Ways typically form earlier and their growth is suppressed at late times. Mass growth associated to major and minor mergers is also biased early for the Milky Way in comparison to M31, with most accretion occurring 1 - 4 Gyr after the Big Bang, and a relatively quiescent history at later times. 32% of our Milky Ways experienced a Gaia-Enceladus/Sausage (GES)-like merger, while 13% host an LMC-like object at the present day, with 5% having both. In one case, an SMC- and a Sagittarius-analogue are also present, showing that the most important mergers of the Milky Way in its Local Group environment can be reproduced in $Λ$CDM. We find that the material that makes up the Milky Way and M31 halos at the present day first collapsed onto a plane roughly aligned with the Local Sheet and Supergalactic plane; after $z \sim 2$, accretion occurred mostly within this plane, with the tidal effects of the heavier companion, M31, significantly impacting the late growth history of the Milky Way.
Bogusław Wacławik, Joanna Popławska, Arkadiusz Sułek
et al.
ESG reporting is a key process through which companies provide detailed information regarding their impact on the environment (E), society (S) and corporate governance (G). The aim of the article is to present the authors' research results on the reporting of environmental information in the field of ESG among companies from the energy sector and the defence industry, represented by the example of PIT-RADWAR, listed on the Warsaw Stock Exchange in 2022-2023. The methodology was based on a review of the literature on the subject, legal acts and own research, which used non-financial reports of listed companies regarding ESG information. According to the authors, the article adds value to the literature on the subject, in particular in terms of gathering source material and discussing it. Considering the growing importance of ESG factors in the defence sector, the article also highlights the emerging trend of ESG reporting in this industry using the example of PIT-RADWAR S.A., emphasising its growing importance and the sector's first steps towards increasing the transparency of non-financial reporting. The subject matter of the article can form the basis for further detailed empirical research on ESG reporting.
With the recent emergence of powerful instruction-tuned large language models (LLMs), various helpful conversational Artificial Intelligence (AI) systems have been deployed across many applications. When prompted by users, these AI systems successfully perform a wide range of tasks as part of a conversation. To provide some sort of memory and context, such approaches typically condition their output on the entire conversational history. Although this sensitivity to the conversational history can often lead to improved performance on subsequent tasks, we find that performance can in fact also be negatively impacted, if there is a task-switch. To the best of our knowledge, our work makes the first attempt to formalize the study of such vulnerabilities and interference of tasks in conversational LLMs caused by task-switches in the conversational history. Our experiments across 5 datasets with 15 task switches using popular LLMs reveal that many of the task-switches can lead to significant performance degradation.
This article offers a new perspective on children and youth’s coerced displacement in the context of the Ottoman Middle East and highlights their potential as a social group to inform studies of children, kinship, and family vis-a-vis the state. Using iltizām contracts, I argue that the Ottoman state prioritized its stability and economic interests and turned a blind eye to promises it made to ensure the “well-being” of young Ottoman subjects. The contracts recorded around the mid-eighteenth century document an institutionalized practice by the state to remove and incarcerate young and minor males associated with the families of multazims, or tax farmers, who generally hailed from the class of provincial notables, or a‘yān, to persuade the latter to render payment of taxes. Although multazims appeared to be indifferent to the fate of their castaway children, evidence suggests that multazims took advantage of geopolitical changes toward the last quarter of the eighteenth century to avoid the incarceration of their children, as the practice completely disappeared at that time. This article also attempts to approach the question of whether this forced displacement of children represents a form of mobility, in comparison to other forms of children’s mobility, like the devširme, and explores what this meant for the expansion, or retraction, of the state power and its governing policies.
This article examines the importance of graphic representations in the social sciences, and particularly in (medieval) history, taking as its starting point a reflection by {É}tienne-Jules Marey, a physiologist and pioneer of 19th-century photography and cinema. Marey believed that the visual should replace language in many fields. Indeed, the twentieth and early twenty-first centuries saw an exponential multiplication of visual media, particularly with the advent of digital technology. However, this ''graphics revolution'' has not affected all disciplines equally. Significant differences remain between scientific fields such as astrophysics, anthropology, chemistry and medieval history, despite their shared commitment to describing dynamic processes and changes of state. Yet, while historians have already digitized a large part of the cultural heritage from Antiquity to the 10th-13th centuries, exploration of this corpus using visualizations remains limited. There is therefore untapped potential in this field.This article begins by outlining a typology and quantification of the past and potential roles of visual representations in medieval history. It examines two distinct intellectual approaches: 1. the use of visuals to support a scientific discourse (majority) and 2. the construction of a historical discourse based on observations made from visual figures with the aim of modeling phenomena invisible to the naked eye. The author thus examines the use of ''images'' in medievalism, focusing on the annual volumes of the Soci{é}t{é} des historiens m{é}di{é}vistes de l'enseignement sup{é}rieur (SHMESP), up to 2006. Two other parts of the text look at the still-rare forms of visual representation in medieval history, particularly those with a ''heuristic vocation'', using iconographic objects, parchments, buildings and digitized texts. The article suggests various visualization techniques, such as network analysis, the creation of ''stemmas 2.0'' and interactive chronologies, which could benefit the discipline. These methods could potentially profoundly change our understanding of ancient societies, by showing the dynamic relationships between different aspects of these societies. One of the most important advances expected from these visual methods is a better understanding of the patterns of development in medieval Europe, which varied from region to region. The hypothesis is that the scarcity of heuristic graphics in medieval history stems from the relationship with ancient documents and the historical method based on narration and exemplarity. The article thus questions the value of ''visual modelling'' in medieval history, and highlights the challenges associated with the widespread adoption of this approach in the humanities and social sciences. Finally, the text invites us to reflect on the nature and functioning of heuristic visual devices, by comparing medieval ''images'' and contemporary scientific visuals. In both cases, the point is to materialize the invisible in order to show something that exists beyond the visual. The author suggests that this way of approaching visuals could play a growing role in the decades to come, particularly in the field of data science.
We introduce Dynamic Contextual Markov Decision Processes (DCMDPs), a novel reinforcement learning framework for history-dependent environments that generalizes the contextual MDP framework to handle non-Markov environments, where contexts change over time. We consider special cases of the model, with a focus on logistic DCMDPs, which break the exponential dependence on history length by leveraging aggregation functions to determine context transitions. This special structure allows us to derive an upper-confidence-bound style algorithm for which we establish regret bounds. Motivated by our theoretical results, we introduce a practical model-based algorithm for logistic DCMDPs that plans in a latent space and uses optimism over history-dependent features. We demonstrate the efficacy of our approach on a recommendation task (using MovieLens data) where user behavior dynamics evolve in response to recommendations.