Search

Goal-conditioned reinforcement learning is a powerful way to control an AI agent's behavior at runtime. That said, popular goal representations, e.g., target states or natural language, are either limited to Markovian tasks or rely on ambiguous task semantics. We propose representing temporal goals using compositions of deterministic finite automata (cDFAs) and use cDFAs to guide RL agents. cDFAs balance the need for formal temporal semantics with ease of interpretation: if one can understand a flow chart, one can understand a cDFA. On the other hand, cDFAs form a countably infinite concept class with Boolean semantics, and subtle changes to the automaton can result in very different tasks, making them difficult to condition agent behavior on. To address this, we observe that all paths through a DFA correspond to a series of reach-avoid tasks and propose pre-training graph neural network embeddings on "reach-avoid derived" DFAs. Through empirical evaluation, we demonstrate that the proposed pre-training method enables zero-shot generalization to various cDFA task classes and accelerated policy specialization without the myopic suboptimality of hierarchical methods.

Background. Fully automatic analysis of myocardial perfusion MRI datasets enables rapid and objective reporting of stress/rest studies in patients with suspected ischemic heart disease. Developing deep learning techniques that can analyze multi-center datasets despite limited training data and variations in software and hardware is an ongoing challenge. Methods. Datasets from 3 medical centers acquired at 3T (n = 150 subjects) were included: an internal dataset (inD; n = 95) and two external datasets (exDs; n = 55) used for evaluating the robustness of the trained deep neural network (DNN) models against differences in pulse sequence (exD-1) and scanner vendor (exD-2). A subset of inD (n = 85) was used for training/validation of a pool of DNNs for segmentation, all using the same spatiotemporal U-Net architecture and hyperparameters but with different parameter initializations. We employed a space-time sliding-patch analysis approach that automatically yields a pixel-wise "uncertainty map" as a byproduct of the segmentation process. In our approach, a given test case is segmented by all members of the DNN pool and the resulting uncertainty maps are leveraged to automatically select the "best" one among the pool of solutions. Results. The proposed DAUGS analysis approach performed similarly to the established approach on the internal dataset (p = n.s.) whereas it significantly outperformed on the external datasets (p < 0.005 for exD-1 and exD-2). Moreover, the number of image series with "failed" segmentation was significantly lower for the proposed vs. the established approach (4.3% vs. 17.1%, p < 0.0005). Conclusions. The proposed DAUGS analysis approach has the potential to improve the robustness of deep learning methods for segmentation of multi-center stress perfusion datasets with variations in the choice of pulse sequence, site location or scanner vendor., Comment: Accepted for publication in JCMR, 2024

This paper investigates the geometric structures and properties of 8-dimensional manifolds with Spin(7)-holonomy. We focus on the characterization and implications of 4-planes within these manifolds, which are endowed with an almost symplectic structure compatible with the Spin(7)-structure. We provide a detailed analysis of the differential forms defining these structures, including the Cayley 4-form and its stabilization under the Spin(7) group actions. Furthermore, we explore the concept of mirror duality within the context of exceptional holonomy, elucidating the relationship between dual structures and their topological constraints. By examining the conditions under which mirror duality arises, we enhance the understanding of the interplay between Spin(7)-structures and symplectic geometry. This endeavor contributes to our deeper understanding of the intricate symmetries and topological properties of these remarkable manifolds., Comment: 14 pages, 1 figure

The binary star NY Hya is a bright, detached, double-lined eclipsing system with an orbital period of just under five days with two components each nearly identical to the Sun and located in the solar neighbourhood. The objective of this study is to test and confront various stellar evolution models for solar-type stars based on accurate measurements of stellar mass and radius. We present new ground-based spectroscopic and photometric as well as high-precision space-based photometric and astrometric data from which we derive orbital as well as physical properties of the components via the method of least-squares minimisation based on a standard binary model valid for two detached components. Classic statistical techniques were invoked to test the significance of model parameters. Additional empirical evidence was compiled from the public domain; the derived system properties were compared with archival broad-band photometry data enabling a measurement of the system's spectral energy distribution that allowed an independent estimate of stellar properties. We also utilised semi-empirical calibration methods to derive atmospheric properties from Str\"{o}mgren photometry and related colour indices. Data was used to confront the observed physical properties with classic and magnetic stellar evolution models., Comment: 34 pages, 19 figures, 13 tables, (accepted for publication in A&A)

We update the ephemerides of 16 transiting exoplanets using our ground-based observations, new TESS data, and previously published observations including those of amateur astronomers. All these light curves were modeled by making use of a set of quantitative criteria with the EXOFAST code to obtain mid-transit times. We searched for statistically significant secular and/or periodic trends in the mid-transit times. We found that the timing data are well modeled by a linear ephemeris for all systems except for XO-2 b, for which we detect an orbital decay with the rate of -12.95 $\pm$ 1.85 ms/yr that can be confirmed with future observations. We also detect a hint of potential periodic variations in the TTV data of HAT-P-13 b which also requires confirmation with further precise observations., Comment: 23 pages, 17 figures

This article reveals a significant connection in geometry: when the Lee form $\theta$ is normal to an almost Hermitian manifold $N$, it implies that $N$ possesses a nearly K\"ahler structure. Investigating locally conformally Spin(7) manifolds with 2-vector fields, our study provides a concise yet rigorous proof of this relationship.

This paper provides an example of structural proxies for black box rings encrypting rings of 2 by 2 matrices of finite fields of odd order.

We have applied a machine learning algorithm to predict the emergence of environment-induced spontaneous synchronization between two qubits in an open system setting. In particular, we have considered three different models, encompassing global and local dissipation regimes, to describe the open system dynamics of the qubits. We have utilized the $k$-nearest neighbors algorithm to estimate the long time synchronization behavior of the qubits only using the early time expectation values of qubit observables in these three distinct models. Our findings clearly demonstrate the possibility of determining the occurrence of different synchronization phenomena with high precision even at the early stages of the dynamics using a machine learning-based approach. Moreover, we show the robustness of our approach against potential measurement errors in experiments by considering random errors in qubit expectation values. We believe that the presented results can prove to be useful in experimental studies on the determination of quantum synchronization., Comment: 13 pages, 10 figures

Dynamic contrast-enhanced (DCE) cardiac magnetic resonance imaging (CMRI) is a widely used modality for diagnosing myocardial blood flow (perfusion) abnormalities. During a typical free-breathing DCE-CMRI scan, close to 300 time-resolved images of myocardial perfusion are acquired at various contrast "wash in/out" phases. Manual segmentation of myocardial contours in each time-frame of a DCE image series can be tedious and time-consuming, particularly when non-rigid motion correction has failed or is unavailable. While deep neural networks (DNNs) have shown promise for analyzing DCE-CMRI datasets, a "dynamic quality control" (dQC) technique for reliably detecting failed segmentations is lacking. Here we propose a new space-time uncertainty metric as a dQC tool for DNN-based segmentation of free-breathing DCE-CMRI datasets by validating the proposed metric on an external dataset and establishing a human-in-the-loop framework to improve the segmentation results. In the proposed approach, we referred the top 10% most uncertain segmentations as detected by our dQC tool to the human expert for refinement. This approach resulted in a significant increase in the Dice score (p<0.001) and a notable decrease in the number of images with failed segmentation (16.2% to 11.3%) whereas the alternative approach of randomly selecting the same number of segmentations for human referral did not achieve any significant improvement. Our results suggest that the proposed dQC framework has the potential to accurately identify poor-quality segmentations and may enable efficient DNN-based analysis of DCE-CMRI in a human-in-the-loop pipeline for clinical interpretation and reporting of dynamic CMRI datasets., Comment: Accepted for publication in MICCAI 2023

The 2023 M7.8 Kahramanmara\c{s}/Pazarcik earthquake was larger and more destructive than what had been expected. Here we analyzed near-field seismic records and developed a dynamic rupture model that reconciles different currently conflicting inversion results and reveals spatially non-uniform propagation speeds in this earthquake, with predominantly supershear speeds observed along the Narli fault and at the southwest (SW) end of the East Anatolian Fault (EAF). The model highlights the critical role of geometric complexity and heterogeneous frictional conditions in facilitating continued propagation and influencing rupture speed. We also constrained the conditions that allowed for the rupture to jump from the Narli fault to EAF and to generate the delayed backpropagating rupture towards the SW. Our findings have important implications for understanding earthquake hazard and guiding future response efforts and demonstrates the value of physics-based dynamic modeling fused with near-field data in enhancing our understanding of earthquake mechanisms and improving risk assessment.

Advancements in simulation and formal methods-guided environment sampling have enabled the rigorous evaluation of machine learning models in a number of safety-critical scenarios, such as autonomous driving. Application of these environment sampling techniques towards improving the learned models themselves has yet to be fully exploited. In this work, we introduce a novel method for improving imitation-learned models in a semantically aware fashion by leveraging specification-guided sampling techniques as a means of aggregating expert data in new environments. Specifically, we create a set of formal specifications as a means of partitioning the space of possible environments into semantically similar regions, and identify elements of this partition where our learned imitation behaves most differently from the expert. We then aggregate expert data on environments in these identified regions, leading to more accurate imitation of the expert's behavior semantics. We instantiate our approach in a series of experiments in the CARLA driving simulator, and demonstrate that our approach leads to models that are more accurate than those learned with other environment sampling methods., Comment: 8 pages, under review

We investigate whether there is a variation in the orbital period of the short-period brown dwarf-mass KELT-1\,b, which is one of the best candidates to observe orbital decay. We obtain 19 high-precision transit light curves of the target using six different telescopes. We add all precise and complete transit light curves from open databases and the literature, as well as the available TESS observations from sectors 17 and 57, to form a transit timing variation (TTV) diagram spanning more than 10 years of observations. The analysis of the TTV diagram, however, is inconclusive in terms of a secular or periodic variation, hinting that the system might have synchronized. We update the transit ephemeris and determine an informative lower limit for the reduced tidal quality parameter of its host star of Q$_{\star}^{\prime} > (8.5 \pm 3.9) \times 10^{6}$ assuming that the stellar rotation is not yet synchronised. Using our new photometric observations, published light curves, the TESS data, archival radial velocities and broadband magnitudes, we also update the measured parameters of the system. Our results are in good agreement with those found in previous analyses., Comment: Accepted for publication by the MNRAS on 2023/01/16

The identification of a deterministic finite automaton (DFA) from labeled examples is a well-studied problem in the literature; however, prior work focuses on the identification of monolithic DFAs. Although monolithic DFAs provide accurate descriptions of systems' behavior, they lack simplicity and interpretability; moreover, they fail to capture sub-tasks realized by the system and introduce inductive biases away from the inherent decomposition of the overall task. In this paper, we present an algorithm for learning conjunctions of DFAs from labeled examples. Our approach extends an existing SAT-based method to systematically enumerate Pareto-optimal candidate solutions. We highlight the utility of our approach by integrating it with a state-of-the-art algorithm for learning DFAs from demonstrations. Our experiments show that the algorithm learns sub-tasks realized by the labeled examples, and it is scalable in the domains of interest., Comment: Preprint, under review

This study introduces a new chemical method for controlling the strain in methylammonium lead iodide (MAPbI3) perovskite crystals by varying the ratio of Pb(Ac)2 and PbCl2 in the precursor solution. To observe the effect on crystal strain, a combination of piezoresponse force microscopy (PFM) and X-ray diffraction (XRD) is used. The PFM images show an increase in the average size of ferroelastic twin domains upon increasing the PbCl2 content, indicating an increase in crystal strain. The XRD spectra support this observation with strong crystal twinning features that appear in the spectra. This behaviour is caused by a strain gradient during the crystallization due to different evaporation rates of methylammonium acetate and methylammonium chloride as revealed by time-of-flight secondary ion mass spectroscopy (ToF-SIMS) and grazing incidince x-ray diffraction (GIXRD) measurements. Additional time-resolved photoluminescence (TRPL) show an increased carrier lifetime in the MAPbI3 films prepared with higher PbCl2 content, suggesting a decreased trap density in films with larger twin domain structures. The results demonstrate the potential of chemical strain engineering as an easy method for controlling strain-related effects in lead halide perovskites.

We study the transit timings of 10 exoplanets in order to investigate potential Transit Timing Variations (TTVs) in them. We model their available ground-based light curves, some presented here and others taken from the literature, and homogeneously measure the mid-transit times. We statistically compare our results with published values and find that the measurement errors agree. However, in terms of recovering the possible frequencies, homogeneous sets can be found to be more useful, of which no statistically relevant example has been found for the planets in our study. We corrected the ephemeris information of all ten planets we studied and provide these most precise light elements as references for future transit observations with space-borne and ground-based instruments. We found no evidence for secular or periodic changes in the orbital periods of the planets in our sample, including the ultra-short period WASP-103 b, whose orbit is expected to decay on an observable timescale. Therefore, we derive the lower limits for the reduced tidal quality factors (Q$^{\prime}_{\star}$) for the host stars based on best fitting quadratic functions to their timing data. We also present a global model of all available data for WASP-74 b, which has a Gaia parallax-based distance value ~25% larger than the published value., Comment: 16 pages, 5 figures, accepted for publication by the MNRAS