Hasil untuk "History"

黄亚明

Max Sobol Mark, Jacky Liang, Maria Attarian et al.

Many robot tasks require attending to the history of past observations. For example, finding an item in a room requires remembering which places have already been searched. However, the best-performing robot policies typically condition only on the current observation, limiting their applicability to such tasks. Naively conditioning on past observations often fails due to spurious correlations: policies latch onto incidental features of training histories that do not generalize to out-of-distribution trajectories upon deployment. We analyze why policies latch onto these spurious correlations and find that this problem stems from limited coverage over the space of possible histories during training, which grows exponentially with horizon. Existing regularization techniques provide inconsistent benefits across tasks, as they do not fundamentally address this coverage problem. Motivated by these findings, we propose Big Picture Policies (BPP), an approach that conditions on a minimal set of meaningful keyframes detected by a vision-language model. By projecting diverse rollouts onto a compact set of task-relevant events, BPP substantially reduces distribution shift between training and deployment, without sacrificing expressivity. We evaluate BPP on four challenging real-world manipulation tasks and three simulation tasks, all requiring history conditioning. BPP achieves 70% higher success rates than the best comparison on real-world evaluations. Videos are available at https://bigpicturepolicies.github.io/

Peter Brimblecombe, Pascal Querner

The insect threat to heritage objects can increase with climate change, increased travel, movement of goods and loan exhibitions. This study used catch from 30 heritage environments across Austria. Overall arthropod catch rate in storerooms was lower than in museums and libraries. Taxonomic richness of the ecosystem in the buildings was a product of building size, perhaps paralleling island biogeography. Heritage pests are distributed independently and follow environmental gradients, perhaps aligning with Henry Gleason’s continuum theory of ecological communities. Catch rates for some abundant pests are evenly distributed among buildings (e.g., <i>Psocoptera</i> booklice, <i>Lepisma saccharinum</i> common silverfish), but <i>Tineola bisselliella</i>, the webbing clothes moth, is unevenly distributed because some locations have large infestations. Rare species are unevenly distributed, as these are found in only a few buildings. A characteristic set of insect pests appear to dominate indoor heritage environments in Austria: <i>Psocoptera</i>, <i>Lepismatidae</i> silverfish, <i>Tineola bisselliella</i> webbing clothes moth and carpet beetles like <i>Anthrenus</i> spp. and <i>Attagenus</i> spp. These pests are also common in the interiors of heritage buildings in some other European countries.

B. Skinner

K. Jaspers

Yuxi Liu, Yanlin Cao, Yugang Li et al.

Focusing on vaccines available to adults and not in the immunization schedule, this study investigates the preferences and factors influencing adults in selecting vaccination clinic locations. It aims to provide strategic insights for boosting vaccination rates by analyzing adults’ decision-making factors. This contributes to developing more efficient, patient-focused vaccination strategies that tackle vaccine hesitancy and improve access to vaccination sites. We conducted a cross-sectional study through the “YueMiao” platform from November 1 to December 10, 2023, using convenience and purposive sampling to engage 2014 participants. We collected data via online surveys that included questions about sociodemographic characteristics, sources of vaccination clinic information, clinic satisfaction, and the impact of site selection on vaccination decisions. Our findings reveal that adults’ site preferences for vaccination are influenced by gender, age, income, and vaccination history. Participants showed a strong preference for locations that offer convenience, efficiency, transparent pricing, and a comfortable environment. Analysis of service satisfaction at these clinics indicates that vaccinated individuals report higher satisfaction with appointment systems, wait times, and service hours than those unvaccinated. Furthermore, the preference for vaccination sites consistently aligns with the vaccine type, with a majority opting for community health service centers. Our results suggest that public health strategies should concentrate on enhancing site convenience, service quality, and information transparency to elevate adult vaccination rates. Future initiatives should aim to increase public trust in vaccines, improve the selection and quality of vaccination sites, and effectively utilize digital technology for spreading vaccination information.

M. Campbell

Koun Shirai

Glass has long been considered a nonequilibrium material. The primary reason is its history-dependent properties: the obtained properties are not uniquely determined by two state variables alone, namely, temperature and volume, but are affected by the process parameters, such as cooling rates. However, closer observations show that this history dependence is common in solid; in crystal growth, the properties of an obtained crystal are affected by the preparation conditions through defect structures and metallurgical structures. The problem with the previous reasoning of history dependence lies in the lack of appropriate specification of state variables. Without knowledge of the latter, describing thermodynamic states is impossible. The guiding principle to find state variables is provided by the first law of thermodynamics. The state variables of solids have been searched by requiring that the internal energy $U$ is a state function. Detailed information about the abovementioned microstructures is needed to describe the state function $U$. This can be accomplished by specifying the time-averaged positions R_{j} of all atoms comprising the solids. Therefore, R_{j} is a state variable for solids. Defect states, being metastable states, represent equilibrium states within a finite time (relaxation time). However, eternal equilibrium is nonexistent: the perfect crystal is thermodynamically unstable. Equilibrium states can only be considered at the local level. Glass is thus in equilibrium as long as its structure does not change. The relaxation time is controlled by the energy barriers by which a structure is sustained, and this time restriction is intimately related to the definition of state variables. The most important property of state variables is their invariance to time averaging. The time-averaged quantity R_{j} meets this invariance property.

Donato Pirovano

La sostanza onirica è una componente importante della narrazione e della sostanza epifanica che caratterizza la Vita nova. Collocati in punti strategici della storia, tutti convergenti verso il (o dal) kérigma della morte o meglio assunzione al cielo di Beatrice, questi episodi si configurano come premonizioni e come aperture dell’orizzonte narrativo. In questo contributo sono analizzate le due, forse tre visiones in somniis, che si trovano rispettivamente nei paragrafi III, XII e XLII. Nella comune dimensione onirica e nella prefigurazione di qualcosa che avverrà tutte e tre mantengono un carattere enigmatico.

Tesfu Fekensa Tujuba, Roman V. Yakovlev, Aidas Saldaitis et al.

African golden-spotted lappet moths from the genera Haplopacha Aurivillius, 1905 and Dasychirinula Hering, 1926 are reviewed. Antennae, heads and legs are investigated in addition to the traditional comparison of habitus, genitalia, and distribution areas. Two new genera are established and four new species are described as a result: Eudoumbia gen. n. with the type-species Eudoumbia thorogood sp. n. from Angola, Namibia, and Botswana; Auripluvia gen. n. with the type species Auripluvia sophia sp. n. from Ethiopia; Haplopacha mason sp. n. from Tanzania; and Dasychirinula julia sp. n. from Ethiopia. One species is reattributed to the new genus as Eudoumbia ndoumoi (Dupont, Simonsen & Zilli, 2016) stat. n. One new synonymy is established for two Tanzanian species: Dasychirinula chrysogramma Hering, 1926 = Haplopacha lunata Dupont, Simonsen & Zilli, 2016 syn. n.; and one potential synonymy is remarked between Haplopacha tangani Dupont, Simonsen & Zilli, 2016 from Tanzania and Malawi and Haplopacha riftensis Dupont, Simonsen & Zilli, 2016 from Malawi. Photos of the related material stored in the main collections of Ditsong National Museum of Natural History (Pretoria, RSA) and Natural History Museum of Zimbabwe (Bulawayo, Zimbabwe) are showed.

Gabriel Brondino, Francisco Leiva, Hernán Alejandro Roitbarg

El trabajo desarrolla una estrategia de estimación no muestral de matriz de coeficientes insumo-producto intrarregionales de escala metropolitana, tomando como caso el Gran Santa Fe. La estrategia metodológica consiste en “regionalizar” la matriz de coeficientes técnicos nacionales. Las fuentes de información utilizadas son la tabla insumo-producto de Argentina de 2018 estimada por la OCDE y datos de empleo recabados por el INDEC. A partir de la estimación, se caracteriza la estructura productiva del GSF en términos de encadenamientos productivos y efectos multiplicadores. El análisis reconstruye el paisaje económico de la región, identificando sus sectores estratégicos, impulsores, clave e independientes

J. Allan

Hao Sun, Yang Li, Liwei Deng et al.

Context information modeling is an important task in conversational KBQA. However, existing methods usually assume the independence of utterances and model them in isolation. In this paper, we propose a History Semantic Graph Enhanced KBQA model (HSGE) that is able to effectively model long-range semantic dependencies in conversation history while maintaining low computational cost. The framework incorporates a context-aware encoder, which employs a dynamic memory decay mechanism and models context at different levels of granularity. We evaluate HSGE on a widely used benchmark dataset for complex sequential question answering. Experimental results demonstrate that it outperforms existing baselines averaged on all question types.

Isobel Falconer

The paper is based on a talk given to celebrate Ørsted's discovery of electromagnetism in 1820. It gives a brief account of Ørsted's life and work, before using the evidence of articles on electricity and magnetism in the Encyclopaedia Britannica to assess understanding of the connection between the two phenomena in the 20 years before Ørsted's discovery, and the immediate reception of it.

Gandharv Patil, Aditya Mahajan, Doina Precup

Reinforcementlearning(RL)folkloresuggeststhathistory-basedfunctionapproximationmethods,suchas recurrent neural nets or history-based state abstraction, perform better than their memory-less counterparts, due to the fact that function approximation in Markov decision processes (MDP) can be viewed as inducing a Partially observable MDP. However, there has been little formal analysis of such history-based algorithms, as most existing frameworks focus exclusively on memory-less features. In this paper, we introduce a theoretical framework for studying the behaviour of RL algorithms that learn to control an MDP using history-based feature abstraction mappings. Furthermore, we use this framework to design a practical RL algorithm and we numerically evaluate its effectiveness on a set of continuous control tasks.

Andreas Ringwald, Carlos Tamarit

The characteristics of the cosmic microwave background provide circumstantial evidence that the hot radiation-dominated epoch in the early universe was preceded by a period of inflationary expansion. Here, we show how a measurement of the stochastic gravitational wave background can reveal the cosmic history and the physical conditions during inflation, subsequent pre- and re-heating, and the beginning of the hot big bang era. This is exemplified with a particularly well-motivated and predictive minimal extension of the Standard Model which is known to provide a complete model for particle physics -- up to the Planck scale, and for cosmology -- back to inflation.

Hina Sakamoto, Kyungjin Ahn, Kiyotomo Ichiki et al.

Cosmic Reionization imprints its signature on the temperature and polarization anisotropies of the cosmic microwave background (CMB). Advances in CMB telescopes have already placed a significant constraint on the history of reionization. As near-future CMB telescopes target the maximum sensitivity, or observations limited only by the cosmic variance (CV), we hereby forecast the potential of future CMB observations in constraining the history of reionization. In this study, we perform Markov Chain Monte Carlo analysis for CV-limited E-mode polarization observations such as the LiteBIRD (Light satellite for the studies of B-mode polarization and Inflation from cosmic background Radiation Detection), based on a few different methods that vary in the way of sampling reionization histories. We focus especially on estimating the very early history of reionization that occurs at redshifts $z>15$, which is quantified by the partial CMB optical depth due to free electrons at $z>15$, $τ_{z>15}$. We find that reionization with $τ_{z>15} \sim 0.008$, which are well below the current upper limit $τ_{z>15} \sim 0.02$, are achievable by reionization models with minihalo domination in the early phase and can be distinguished from those with $τ_{z>15} \lesssim 5\times 10^{-4}$ through CV-limited CMB polarization observations. An accurate estimation of $τ_{z>15}$, however, remains somewhat elusive. We investigate whether resampling the E-mode polarization data with limited spherical-harmonic modes may resolve this shortcoming.

Tae-hyeon Shin, Benedikt Diemer

The density profiles of dark matter haloes contain rich information about their growth history and physical properties. One particularly interesting region is the splashback radius, $R_{\rm sp}$, which marks the transition between particles orbiting in the halo and particles undergoing first infall. While the dependence of $R_{\rm sp}$ on the recent accretion rate is well established and theoretically expected, it is not clear exactly what parts of the accretion history $R_{\rm sp}$ responds to, and what other halo properties might additionally influence its position. We comprehensively investigate these questions by correlating the dynamically measured splashback radii of a large set of simulated haloes with their individual growth histories as well as their structural, dynamical, and environmental properties. We find that $R_{\rm sp}$ is sensitive to the accretion over one crossing time but largely insensitive to the prior history (in contrast to concentration, which probes earlier epochs). All secondary correlations are much weaker, but we discern a relatively higher $R_{\rm sp}$ in less massive, older, more elliptical, and more tidally deformed haloes. Despite these minor influences, we conclude that the splashback radius is a clean indicator of a halo's growth over the past dynamical time. We predict that the magnitude gap should be a promising observable indicator of a halo's accretion rate and splashback radius.

Guoping Rong, Yifan Zhang, Lanxin Yang et al.

Modern code review is a critical and indispensable practice in a pull-request development paradigm that prevails in Open Source Software (OSS) development. Finding a suitable reviewer in projects with massive participants thus becomes an increasingly challenging task. Many reviewer recommendation approaches (recommenders) have been developed to support this task which apply a similar strategy, i.e. modeling the review history first then followed by predicting/recommending a reviewer based on the model. Apparently, the better the model reflects the reality in review history, the higher recommender's performance we may expect. However, one typical scenario in a pull-request development paradigm, i.e. one Pull-Request (PR) (such as a revision or addition submitted by a contributor) may have multiple reviewers and they may impact each other through publicly posted comments, has not been modeled well in existing recommenders. We adopted the hypergraph technique to model this high-order relationship (i.e. one PR with multiple reviewers herein) and developed a new recommender, namely HGRec, which is evaluated by 12 OSS projects with more than 87K PRs, 680K comments in terms of accuracy and recommendation distribution. The results indicate that HGRec outperforms the state-of-the-art recommenders on recommendation accuracy. Besides, among the top three accurate recommenders, HGRec is more likely to recommend a diversity of reviewers, which can help to relieve the core reviewers' workload congestion issue. Moreover, since HGRec is based on hypergraph, which is a natural and interpretable representation to model review history, it is easy to accommodate more types of entities and realistic relationships in modern code review scenarios. As the first attempt, this study reveals the potentials of hypergraph on advancing the pragmatic solutions for code reviewer recommendation.

Dharma Kumar, T. Raychaudhuri, I. Habib