31325 research outputs found
Sort by
Motor styles in action: Developing a computational framework for operationalization of motor distances
No full textAside from some common movement regularities, significant inter-individual and inter-trial variation within the same individual exists in motor system output. However, there is still a lack of a robust and widely adopted solution for quantifying the degree of similarity between movements. We therefore developed an innovative approach based on the Procrustes transformation to compute 'motor distance' between pairs of kinematic data. As a proof of concept, we tested this on a dataset of reach-to-grasp movements performed by 16 participants while acting with the same confederate. Using the information of wrist velocity, acceleration, and jerk, the proposed technique was able to correctly estimate smaller distances between movements performed by the confederate compared with those of participants. Moreover, the reconstructed pattern of inter-subject distances was consistent when computed either on precision grip prehension or whole hand prehension, suggesting its suitability for the investigation of 'motor styles'. The definition of a solid approach to 'motor distance' computation, therefore, opens the way to new research lines in the field of movement kinematics.Peer Reviewe
Combinatorial refinement on circulant graphs
No full textThe combinatorial refinement techniques have proven to be an efficient approach to isomorphism testing for particular classes of graphs. If the number of refinement rounds is small, this puts the corresponding isomorphism problem in a low-complexity class. We investigate the round complexity of the two-dimensional Weisfeiler--Leman algorithm on circulant graphs, i.e., on Cayley graphs of the cyclic group Zn, and prove that the number of rounds until stabilization is bounded by O(d(n)logn), where d(n)is the number of divisors of n. As a particular consequence, isomorphism can be tested in NC for connected circulant graphs of order pℓwith p an odd prime, ℓ>3and vertex degree Δsmaller than p. We also show that the color refinement method (also known as the one-dimensional Weisfeiler--Leman algorithm) computes a canonical labeling for every non-trivial circulant graph with a prime number of vertices after individualization of two appropriately chosen vertices. Thus, the canonical labeling problem for this class of graphs has at most the same complexity as color refinement, which results in a time bound of O(Δnlogn). Moreover, this provides a first example where a sophisticated approach to isomorphism testing put forward by Tinhofer has a real practical meaning.Peer Reviewe
Science with a Small Two-Band UV-Photometry Mission I: Mission Description and Follow-up Observations of Stellar Transients
No full textThis is the first in a collection of three papers introducing the science with an ultra-violet (UV) space telescope on an approximately 130 kg small satellite with a moderately fast re-pointing capability and a real-time alert communication system approved for a Czech national space mission. The mission, called Quick Ultra-Violet Kilonova surveyor—QUVIK , will provide key follow-up capabilities to increase the discovery potential of gravitational wave observatories and future wide-field multi-wavelength surveys. The primary objective of the mission is the measurement of the UV brightness evolution of kilonovae, resulting from mergers of neutron stars, to distinguish between different explosion scenarios. The mission, which is designed to be complementary to the Ultraviolet Transient Astronomy Satellite—ULTRASAT , will also provide unique follow-up capabilities for other transients both in the near- and far-UV bands. Between the observations of transients, the satellite will target other objects described in this collection of papers, which demonstrates that a small and relatively affordable dedicated UV-space telescope can be transformative for many fields of astrophysics.Open access publishing supported by the National Technical Library in Prague.Masaryk UniversityPeer Reviewe
Valence without meaning: Investigating form and semantic components in pseudowords valence
No full textValence is a dominant semantic dimension, and it is fundamentally linked to basic approach-avoidance behavior within a broad range of contexts. Previous studies have shown that it is possible to approximate the valence of existing words based on several surface-level and semantic components of the stimuli. Parallelly, recent studies have shown that even completely novel and (apparently) meaningless stimuli, like pseudowords, can be informative of meaning based on the information that they carry at the subword level. Here, we aimed to further extend this evidence by investigating whether humans can reliably assign valence to pseudowords and, additionally, to identify the factors explaining such valence judgments. In Experiment 1, we trained several models to predict valence judgments for existing words from their combined form and meaning information. Then, in Experiment 2 and Experiment 3, we extended the results by predicting participants’ valence judgments for pseudowords, using a set of models indexing different (possible) sources of valence and selected the best performing model in a completely data-driven procedure. Results showed that the model including basic surface-level (i.e., letters composing the pseudoword) and orthographic neighbors information performed best, thus tracing back pseudoword valence to these components. These findings support perspectives on the nonarbitrariness of language and provide insights regarding how humans process the valence of novel stimuli.Open access funding provided by Università degli Studi di Pavia within the CRUI-CARE Agreement.Deutsche Forschungsgemeinschafthttp://dx.doi.org/10.13039/501100001659Università degli Studi di PaviaPeer Reviewe
Testing AI on language comprehension tasks reveals insensitivity to underlying meaning
No full textLarge Language Models (LLMs) are recruited in applications that span from clinical assistance and legal support to question answering and education. Their success in specialized tasks has led to the claim that they possess human-like linguistic capabilities related to compositional understanding and reasoning. Yet, reverse-engineering is bound by Moravec’s Paradox, according to which easy skills are hard. We systematically assess 7 state-of-the-art models on a novel benchmark. Models answered a series of comprehension questions, each prompted multiple times in two settings, permitting one-word or open-length replies. Each question targets a short text featuring high-frequency linguistic constructions. To establish a baseline for achieving human-like performance, we tested 400 humans on the same prompts. Based on a dataset of n = 26,680 datapoints, we discovered that LLMs perform at chance accuracy and waver considerably in their answers. Quantitatively, the tested models are outperformed by humans, and qualitatively their answers showcase distinctly non-human errors in language understanding. We interpret this evidence as suggesting that, despite their usefulness in various tasks, current AI models fall short of understanding language in a way that matches humans, and we argue that this may be due to their lack of a compositional operator for regulating grammatical and semantic information.European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No. 945413 and from the Universitat Rovira i VirgiliGerman Research Foundation (Deutsche Forschungsgemeinschaft) under the Emmy-Noether grant “What’s in a name?” (project No. 459717703)Spanish Ministry of Science, Innovation & Universities (MCIN/AEI/10.13039/501100011033)Peer Reviewe
Evaluation of a CBT‐Based Program for Mental Health in the General Population During the COVID‐19 Pandemic: A Stepped‐Care Approach Using a Chatbot and Digitized Group Intervention
No full textBackground: The COVID‐19 pandemic exposed a substantial portion of society to multiple stressors, while access to mental health care was limited. To address this, we introduced a digital stepped‐care program rooted in cognitive–behavioral therapy (CBT) principles, aiming to alleviate mental health distress among the general public seeking help. Methods: The program comprises a 4‐week digital application using “Aury” the chatbot, followed by an optional 6‐week online group session for those still symptomatic. A 4‐week waiting period separated these steps. Participants entered based on self‐identified mental health concerns. Interventions addressed prevalent pandemic mental health issues: sleep disturbances, anxiety, depression, worry/rumination, interpersonal issues, and resource mobilization. Outcomes focused on depressive, anxiety, and somatic symptoms, assessed by the Patient Health Questionnaire (PHQ). Results: Of the 1261 initial participants, postchatbot results ( N = 142) indicated small to medium effects ( d = 0.412 to d = 0.523). Those finishing the entire program ( N = 41) saw substantial symptom decline with medium to large effects ( d = 0.757 to d = 0.818). No shifts were seen in the waiting phase. At follow‐up 6 months after baseline, both groups—those who only used the chatbot ( N = 60; d = 0.284 to d = 0.416) and those who completed the entire program ( N = 27; d = 0.854 to d = 0.926)—showed sustained symptom reduction. Comparing groups that received no intervention, used the chatbot only, and completed the entire program, we observed a dose–response effect. Conclusions: This resource‐efficient and adaptable digital approach effectively reduced pandemic‐induced mental health issues, indicating its potential in crisis periods with limited health resources. Randomized controlled trials are recommended for further validation. Trial Registration: Clinical Trial Registry identifier: DRKS00023220 .Peer Reviewe
π‐Lewis Base Activation of Carbonyls and Hexafluorobenzene
No full textWe report hitherto elusive side‐on η2‐bonded palladium(0) carbonyl (anthraquinone, benzaldehyde) and arene (benzene, hexafluorobenzene) palladium(0) complexes and present the catalytic hydrodefluorination of hexafluorobenzene by cyclohexene. The comparison with respective cyclohexene, pyridine and tetrahydrofuran complexes reveals that the experimental ligand binding strengths follow the order THF<C6H6<C6F6<cyclohexene<pyridine<benzaldehyde<anthraquinone. To understand this surprising order, the complexes’ electronic structures were elucidated by nuclear magnetic resonance (NMR), single crystal X‐Ray diffraction (sc‐XRD), ultraviolet/visible (UV/Vis) electronic absorption, infrared (IR) vibrational, Pd L3‐edge X‐ray absorption (XAS), and X‐ray photoelectron (XP) spectroscopic techniques, complemented by Density Functional Theory (DFT) calculations including energy decomposition (EDA‐NOCV) and effective oxidation state (EOS) analyses. For benzene, pyridine and cyclohexene, bonding follows the donor/acceptor picture of the Dewar–Chatt–Duncanson model. In stark contrast, hexafluorobenzene, benzaldehyde and anthraquinone bind via essentially the π‐channel only and thus as π‐analogues of Z‐acceptor ligands. This contribution elucidates the control of functional‐group selectivity in palladium(0) catalysis and delineates a novel strategy to activate electron‐deficient π‐systems.H2020 European Research Council http://dx.doi.org/10.13039/100010663Deutsche Forschungsgemeinschaft http://dx.doi.org/10.13039/501100001659Peer Reviewe
Learning from small data sets: Patch‐based regularizers in inverse problems for image reconstruction
No full textThe solution of inverse problems is of fundamental interest in medical and astronomical imaging, geophysics as well as engineering and life sciences. Recent advances were made by using methods from machine learning, in particular deep neural networks. Most of these methods require a huge amount of data and computer capacity to train the networks, which often may not be available. Our paper addresses the issue of learning from small data sets by taking patches of very few images into account. We focus on the combination of model‐based and data‐driven methods by approximating just the image prior, also known as regularizer in the variational model. We review two methodically different approaches, namely optimizing the maximum log‐likelihood of the patch distribution, and penalizing Wasserstein‐like discrepancies of whole empirical patch distributions. From the point of view of Bayesian inverse problems, we show how we can achieve uncertainty quantification by approximating the posterior using Langevin Monte Carlo methods. We demonstrate the power of the methods in computed tomography, image super‐resolution, and inpainting. Indeed, the approach provides also high‐quality results in zero‐shot super‐resolution, where only a low‐resolution image is available. The article is accompanied by a GitHub repository containing implementations of all methods as well as data examples so that the reader can get their own insight into the performance.Peer Reviewe
Additively Manufactured Ceramics for Compact Quantum Technologies
No full textQuantum technologies are advancing from fundamental research in specialized laboratories to practical applications in the field, driving the demand for robust, scalable, and reproducible system integration techniques. Ceramic components can be pivotal thanks to high stiffness, low thermal expansion, and excellent dimensional stability under thermal stress. Lithography‐based additive manufacturing of technical ceramics is explored, especially for miniaturized physics packages and electro‐optical systems. This approach enables functional systems with precisely manufactured, intricate structures, and high mechanical stability while minimizing size and weight. It facilitates rapid prototyping, simplifies fabrication and leads to highly integrated, reliable devices. As an electrical insulator with low outgassing and high temperature stability, printed technical ceramics such as Al2O3and AlN bridge a technology gap in quantum technology and offer advantages over other printable materials. This potential is demonstrated with CerAMRef, a micro‐integrated rubidium D2 line optical frequency reference on a printed micro‐optical bench and housing. The frequency instability of the reference is comparable to laboratory setups while the volume of the integrated spectroscopy setup is only 6mL. Potential for future applications is identified in compact atomic magnetometers, miniaturized optical atom traps, and vacuum system integration.Bundesministerium für Wirtschaft und Klimaschutz http://dx.doi.org/10.13039/100021130German Space AgencyFederal Ministry for Economic Affairs and Climate ActionPeer Reviewe
Predicting recurrent chat contact in a psychological intervention for the youth using natural language processing
No full textChat-based counseling hotlines emerged as a promising low-threshold intervention for youth mental health. However, despite the resulting availability of large text corpora, little work has investigated Natural Language Processing (NLP) applications within this setting. Therefore, this preregistered approach (OSF: XA4PN) utilizes a sample of approximately 19,000 children and young adults that received a chat consultation from a 24/7 crisis service in Germany. Around 800,000 messages were used to predict whether chatters would contact the service again, as this would allow the provision of or redirection to additional treatment. We trained an XGBoost Classifier on the words of the anonymized conversations, using repeated cross-validation and bayesian optimization for hyperparameter search. The best model was able to achieve an AUROC score of 0.68 ( p < 0.01) on the previously unseen 3942 newest consultations. A shapely-based explainability approach revealed that words indicating younger age or female gender and terms related to self-harm and suicidal thoughts were associated with a higher chance of recontacting. We conclude that NLP-based predictions of recurrent contact are a promising path toward personalized care at chat hotlines.Peer Reviewe