An increasing body of research has demonstrated that human learners are able to use co-occurrences among wordsand objects to form word-object associations (e.g., Yu & Smith, 2007). In this study, we further investigated learners’ ability touse statistical information to learn labels at different hierarchical levels. Participants were presented with objects and words inambiguous learning trials. In some learning trials, participants saw multiple objects and heard their individual labels presentedin a random order, while in other trials, category labels were presented instead. Results from three experiments providedconverging evidence that adults were able to use word-object co-occurrences across different situations to learn hierarchicallabels. Moreover, participants generalized category labels to novel members at the same level but not to superordinate-levelinstances. There was also an interaction between the level of ambiguity in learning contexts and performance in label learningand generalization.

Satellite observations show that the surface urban heat island intensity (SUHII) has been increasing over the last two decades. This is often accompanied by an increased urban-rural contrast of vegetation greenness. However, the contribution of uneven vegetation trends in urban and rural areas to the trend of SUHII is unclear, due to the confounding effects of climate change and changes in man-made infrastructures and anthropogenic heat sources. Here we use a data-model fusion approach to quantify such contributions during the peak growing season. We show that the LAIdif (the urban-rural difference of leaf area index) is increasing (P< 0.05) in 189 of the selected 228 global megacities. The increasing trend of LAIdif from 2000 to 2019 accounts for about one quarter of the trend in satellite-derived SUHII, and the impact is particularly evident in places with rapid urbanization and rural cropland intensification. The marginal sensitivity of SUHII to LAIdif is the strongest in hot-humid megacities surrounded by croplands and in hot-dry megacities surrounded by mixed woody and herbaceous vegetation. Our study highlights the role of long-term vegetation trends in modulating the trends of urban-rural temperature differences.

The present study investigates whether, and if so in what way, adult learners are sensitive to the properties ofthe statistical input, such as frequency and skewedness, when learning and generalizing category labels. Participants werepresented with novel objects belonging to four different categories and heard category labels in a cross-situational learningtask. The four categories were matched for the total amount of exposure but varied in category size and shape of distribution.Participants learned object-to-label mappings better for categories with a skewed distribution of fewer objects. Moreover,object-to-label mapping performance was positively related to the ability to extend category knowledge to novel items. Co-occurrence frequency or category size alone were not good predictors of label learning and generalization. The results indicatethe importance of input distribution in word and category learning processes.

Crosslevel is defined as the difference in elevation between the top surface of two railroad tracks. Severe changes in crosslevel, for example, due to earthquakes, ground settlement, or crushed ballasts, affect track geometry and can cause train derailment. Therefore, the objective of this study was to monitoring railroad crosslevel by using electrical time domain reflectometry (ETDR) to simultaneously interrogate multiple capacitive tilt sensor prototypes connected in a transmission line. ETDR works by propagating an electrical pulse signal from one end of the transmission line and then monitoring the characteristics of each reflected pulse, which is affected by the capacitance (or tilt) of the sensors. This study begins with a discussion of the capacitive tilt sensor's design. These 3D-printed sensors were tested to characterize their tilt sensing performance. Then, multiple tilt sensors were connected in a transmission line and interrogated by ETDR. The ability to use ETDR to multiplex and interrogate sensors subjected to different angles of tilt was validated.

Machine-learned interatomic potentials based on local environment descriptors represent a transformative leap over traditional potentials based on rigid functional forms in terms of prediction accuracy. However, a challenge in their application to ionic systems is the treatment of long-ranged electrostatics. Here, we present a highly accurate electrostatic Spectral Neighbor Analysis Potential (eSNAP) for ionic α-Li3N, a prototypical lithium superionic conductor of interest as a solid electrolyte or coating for rechargeable lithium-ion batteries. We show that the optimized eSNAP model substantially outperforms traditional Coulomb–Buckingham potential in the prediction of energies and forces, as well as various properties, such as lattice constants, elastic constants, and phonon dispersion curves. We also demonstrate the application of eSNAP in long-time, large-scale Li diffusion studies in Li3N, providing atomistic insights into measures of concerted ionic motion (e.g., the Haven ratio) and grain boundary diffusion. This work aims at providing an approach to developing quantum-accurate force fields for multi-component ionic systems under the SNAP formalism, enabling large-scale atomistic simulations for such systems.

Analyzing coordination environments using X-ray absorption spectroscopy has broad applications in solid-state physics and material chemistry. Here, we show that random forest models trained on 190,000 K-edge X-ray absorption near-edge structure (XANES) spectra can identify the main atomic coordination environment with a high accuracy of 85.4% and all associated coordination environments with a high Jaccard score of 81.8% for 33 cation elements in oxides, significantly outperforming other machine-learning models. In a departure from prior works, the coordination environment is described as a distribution over 25 distinct coordination motifs with coordination numbers ranging from 1 to 12. More importantly, we show that the random forest models can be used to predict coordination environments from experimental K-edge XANES with minimal loss in accuracy. A drop-variable feature importance analysis highlights the key roles that the pre-edge and main-peak regions play in coordination environment identification.

The ability to predict others actions is fundamental to successful joint action, communication, and theory of mind. Re-search has shown that infants predict other peoples actions across a variety of laboratory tasks. However, it is unknownwhether the action prediction skills that infants demonstrate during screen-based eye-tracking tasks scale up to real-lifeaction contexts, and whether they relate to general learning abilities. To address these questions, we used head-mountedeye-tracking to investigate action prediction and visual sequence learning during live parent-child interactions. Findingsreveal that 18-month-old infants predict reaching actions during the majority of trials, and that their gaze latencies becomefaster as they learn 3-step action sequences. These findings demonstrate that infants can learn sequence regularities andanticipate the actions of other people in live, naturalistic contexts, as they have been shown to do in traditional laboratorycontexts. This research contributes new insight into early cognitive and social development.

Developmental theory considers action prediction as one of several processes involved in determining how infants come to perceive and understand social events (Gredebäck & Daum, 2015). Action prediction is observed from early in life and is considered an important social-cognitive skill. However, knowledge about infant action prediction is limited to evidence from screen-based eye-tracking tasks. Little is known about action prediction in real-life action contexts. Our aim in the current study was to provide new evidence on whether and how infants anticipate actions in free-flowing parent-child interaction. Using dual head-mounted eye- tracking, we analyzed infants’ visual anticipations of their parents’ reaching actions while they played with objects together. Findings reveal that infants anticipate their parents’ actions at a rate higher than would be expected by chance.