Search

Teacher shortages and lack of teacher diversity have led to growing efforts nationally to recruit teaching assistants (TAs) to be classroom teachers. Substitute teachers are not typically considered in these efforts. We pair longitudinal administrative data from a mid-sized urban district with survey follow-up to address how TAs and substitute teachers contribute to filling staffing shortages and diversifying the teacher workforce. While substitute teachers are rarely included in formal Grow-Your-Own efforts, they bring racial and ethnic diversity to the district in similar ways to TAs yet face fewer barriers to becoming classroom teachers. As a result, they do so at much higher rates and help prevent vacancies in hard-to-staff classrooms.

Leaders in higher education face increasing pressure to ensure their institutions are well-positioned to adapt to our changing world. As part of responding to these demands, a growing number of institutions have established dedicated teams of in-house experts to support this work and the culture change around it, broadly termed academic change or academic innovation. This report details methods and findings of a comprehensive survey of academic innovation department leaders in colleges and universities across the United States. The survey received responses from 138 academic innovation leaders who shared details on the scope of their work, the structures and institutional resources supporting it, their impact, and their perspectives on contemporary challenges including the COVID-19 pandemic and its influence on postsecondary learning. Ten years after the first such survey, findings included evidence of increasing staffing and budgets, evolving priorities, shifting reporting structures, and technological advancement. Implications include expanding research and professional communities for this relatively new functional area within US higher education, recommendations for institutional leaders, and a focus on supporting continued growth. As a result, the report serves as a census of academic innovation units in an array of US colleges and universities, with details relevant for benchmarking and further research. The following are appended: (1) Data tables for all survey items; (2) Participating institution list; and (3) Survey instrument. [This report was produced by the University of Michigan, Center for Academic Innovation and Quantum Thinking. Additional sponsors for this report include: Class Technologies, Acadeum, Intelliboard, Auburn University, Biggio Center for the Enhancement of Teaching and Learning Leading Academic Change, EdPlus at Arizona State University, Bentley University, University System of Maryland, William E. Kirwan Center for Academic Innovation, POD Network in Higher education (POD), and SAB Creative & Consulting.]

We consider rotational beta expansions in dimensions 1, 2 and 4 and view them as expansions on real numbers, complex numbers, and quaternions, respectively. We give sufficient conditions on the parameters $\alpha, \beta \in (0,1)$ so that particular cylinder sets arising from the expansions are winning or losing Schmidt $(\alpha,\beta)$-game.

While high-resolution microscopic techniques are crucial for studying cellular structures in cell biology, obtaining such images from thick 3D engineered tissues remains challenging. In this review, we explore advancements in fluorescence microscopy, alongside the use of various fluorescent probes and material processing techniques to address these challenges. We navigate through the diverse array of imaging options available in tissue engineering field, from wide field to super-resolution microscopy, so researchers can make more informed decisions based on the specific tissue and cellular structures of interest. Finally, we provide some recent examples of how traditional limitations on obtaining high-resolution images on sub-cellular architecture within 3D tissues have been overcome by combining imaging advancements with innovative tissue engineering approaches.

We find new estimates and a new asymptotic decoupling phenomenon for solutions to Hitchin's self-duality equations at high energy. These generalize previous results for generically regular semisimple Higgs bundles to arbitrary Higgs bundles. We apply our estimates to the Hitchin WKB problem and to high energy harmonic maps to symmetric spaces and buildings.

Particularly for aneutronic fusion schemes, it is advantageous to manipulate the fuel species differently from one another, as well as expel ash promptly. The ponderomotive effect can be used to selectively manipulate particles. It is commonly a result of particle-wave interactions and has a complex dependence on the particle charge and mass, enabling species-selectivity. If the plasma is rotating, e.g. due to $\mathbf{E} \times \mathbf{B}$ motion, the ponderomotive effect can be generated using static (i.e., time-independent) perturbations to the electric and magnetic fields, which can be significantly cheaper to produce than time-dependent waves. This feature can be particularly useful in rotating mirror machines where mirror confinement can be enhanced by rotation, both through centrifugal confinement and additionally through a ponderomotive interaction with a static azimuthal perturbation. Some static perturbations generate a ponderomotive barrier, other perturbations can generate either a repulsive barrier or an attractive ponderomotive well which can be used to attract particles of a certain species while repelling another. The viability of each of these effects depends on the specifics of the rotation profile and temperature, and the resultant dispersion relation in the rotating plasma., Comment: 11 figures, 12 pages, Papers from the 66th Annual Meeting of the APS Division of Plasma Physics

Bad actors in the maritime industry engage in illegal behaviors after disabling their vessel's automatic identification system (AIS) - which makes finding such vessels difficult for analysts. Machine learning approaches only succeed in identifying the locations of these ``dark vessels'' in the immediate future. This work leverages ideas from the literature on abductive inference applied to locating adversarial agents to solve the problem. Specifically, we combine concepts from abduction, logic programming, and rule learning to create an efficient method that approaches full recall of dark vessels while requiring less search area than machine learning methods. We provide a logic-based paradigm for reasoning about maritime vessels, an abductive inference query method, an automatically extracted rule-based behavior model methodology, and a thorough suite of experiments., Comment: Accepted to 24th International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2025)

Handling long-range dependencies in neural architectures has remained a persistent challenge due to computational limitations and inefficient contextual retention mechanisms. Tensorial operations have provided a foundation for restructuring model representations, yet conventional architectures have struggled to incorporate such techniques without introducing excessive complexity. A novel approach, Context-Preserving Tensorial Reconfiguration (CPTR), enables dynamic reorganization of weight tensors through structured factorization and adaptive contraction, allowing for enhanced contextual integration without substantial computational overhead. Empirical evaluations demonstrate that CPTR improves coherence retention across extended sequences, leading to measurable reductions in perplexity and improved recall accuracy for long-context tasks. Performance comparisons reveal that CPTR-enhanced models exhibit greater computational efficiency and reduced memory consumption while maintaining competitive language generation fluency and accuracy. Gradient stability metrics further validate the improved training efficiency, revealing more controlled variance in weight updates. Comparative studies across baseline and CPTR-enhanced models confirm that tensorial reconfiguration contributes to more stable and computationally efficient language modeling. The findings support the potential of CPTR in refining contemporary neural architectures for tasks requiring long-range contextual understanding and efficient memory utilization.

In this paper an extensive database of SPARC H-modes confinement predictions has been provided, to assess its variability with respect to few input assumptions. The simulations have been performed within the ASTRA framework, using the quasi-linear model TGLF SAT2, including electromagnetic effects, for the core transport, and a neural network trained on EPED simulations to predict the pedestal height and width self-consistently. The database has been developed starting from two SPARC H-mode discharges (12.2 T, i.e. Primary Reference Discharge or PRD, and 8 T, i.e. reduced field) and permuting 4 input parameters (W concentration, DT mixture concentration, temperature ratio at top of pedestal and deviation of pedestal pressure from the EPED prediction), to perform a sensitivity study. For the PRD a scan of auxiliary input power (ion cyclotron heating) has been performed up to 25MW, to keep highly radiative plasmas above the LH power threshold as predicted by Martin and Schmidtmayr power scalings. A scan of pedestal density has then been performed for both PRD and 8T databases. ptop/pEPED and Ti/Te at top of pedestal showed the biggest impact on the fusion gain. Significant variation is observed across the database, highlighting the importance of sensitivity studies. Below a certain W concentration, the 12T database shows that Q > 5 is consistently achieved for full-field H-modes with 11 MW of auxiliary power, and values of Q > 2 are assured when increasing the input power to keep the plasma in H-mode. The 8T database demonstrates that SPARC can access a Q > 1 operational window with low W concentration, making it a potentially interesting scenario for obtaining breakeven conditions.

Transformers have demonstrated great success in numerous domains including natural language processing and bioinformatics. This success stems from the use of the attention mechanism by these models in order to represent and propagate pairwise interactions between individual tokens of sequential data. However, the primary limitation of this operation is its quadratic memory and time complexity in relation to the input's context length - the length of a sequence over which the interactions need to be captured. This significantly limits the length of sequences that can be inferred upon by these models. Extensive research has been conducted to reduce the number of pairwise interactions to sub-quadratic in relation to the context length by introducing sparsity into the attention mechanism through the development of sparse attention masks. However, efficient implementations that achieve "true sparsity" are lacking. In this work, we address this issue by proposing a graph computing view of attention where tokens are perceived as nodes of the graph and the attention mask determines the edges of the graph. Using this view, we develop graph processing algorithms to implement the attention mechanism. Both theoretically and empirically, we demonstrate that our algorithms only perform the needed computations, i.e., they are work optimal. We also perform extensive experimentation using popular attention masks to explore the impact of sparsity on execution time and achievable context length. Our experiments demonstrate significant speedups in execution times compared to state-of-the-art attention implementations such as FlashAttention for large sequence lengths. We also demonstrate that our algorithms are able to achieve extremely long sequence lengths of as high as 160 million on a single NVIDIA A100 GPU (SXM4 80GB).

Encryption-based cyber threats continue to evolve, leveraging increasingly sophisticated cryptographic techniques to evade detection and persist within compromised systems. A hierarchical classification framework designed to analyze structural cryptographic properties provides a novel approach to distinguishing malicious encryption from legitimate cryptographic operations. By systematically decomposing encryption workflows into hierarchical layers, the classification method enhances the ability to recognize distinct patterns across diverse threat variants, reducing the dependence on predefined signatures that often fail against rapidly mutating threats. The study examines how cryptographic feature mapping facilitates improved classification accuracy, highlighting the role of entropy, key exchange mechanisms, and algorithmic dependencies in distinguishing harmful encryption activities. Through experimental validation, the framework demonstrated a high degree of precision across multiple attack families, outperforming conventional classification techniques while maintaining computational efficiency suitable for large-scale cybersecurity applications. The layered structural analysis further enhances forensic investigations, enabling security analysts to dissect encryption workflows to trace attack origins and identify commonalities across different campaigns. The methodology strengthens proactive threat mitigation efforts, offering a scalable and adaptable solution that accounts for both known and emerging encryption-based cyber threats. Comparative evaluations illustrate the advantages of structural decomposition in mitigating false positives and negatives, reinforcing the reliability of cryptographic signature classification in real-world security environments.

Autonomous systems are increasingly expected to operate in the presence of adversaries, though an adversary may infer sensitive information simply by observing a system, without even needing to interact with it. Therefore, in this work we present a deceptive decision-making framework that not only conceals sensitive information, but in fact actively misleads adversaries about it. We model autonomous systems as Markov decision processes, and we consider adversaries that attempt to infer their reward functions using inverse reinforcement learning. To counter such efforts, we present two regularization strategies for policy synthesis problems that actively deceive an adversary about a system's underlying rewards. The first form of deception is ``diversionary'', and it leads an adversary to draw any false conclusion about what the system's reward function is. The second form of deception is ``targeted'', and it leads an adversary to draw a specific false conclusion about what the system's reward function is. We then show how each form of deception can be implemented in policy optimization problems, and we analytically bound the loss in total accumulated reward that is induced by deception. Next, we evaluate these developments in a multi-agent sequential decision-making problem with one real agent and multiple decoys. We show that diversionary deception can cause the adversary to believe that the most important agent is the least important, while attaining a total accumulated reward that is $98.83\%$ of its optimal, non-deceptive value. Similarly, we show that targeted deception can make any decoy appear to be the most important agent, while still attaining a total accumulated reward that is $99.25\%$ of its optimal, non-deceptive value., Comment: 21 pages, 5 figures

Highly reflective Calcium Phosphate (CAP) nanoparticles have been obtained from waste chicken and porcine bones. Chicken and pork bones have been processed and calcined at temperatures between 600{\deg}C and 1200{\deg}C to remove organic material and resulting in CAP bio-ceramic compounds with high reflectance. The reflectivity of the materials in the solar wavelength region is on par with chemically synthesized CAP. The high reflectivity, consistently over 90%, as well as the size distribution and packing density of the nanoparticles obtained in these early bone studies make a strong case for pursuing this avenue to obtain pigment for high solar reflectivity applications, such as passive daytime radiative cooling. The results presented indicate a viable path toward a cost-effective and eco-friendly source of highly reflective cooling pigments. By sourcing calcium phosphates from animal bones, there is also the potential to divert large quantities of bone waste generated by the meat industry from landfills, further contributing toward sustainability and energy reduction efforts in the construction industry and beyond., Comment: 15 pages, 4 figures

We compare X-ray emission from several general relativistic, multi-frequency, radiation magnetohydrodynamic simulations of thin black hole accretion disks with different accretion rates and spins. The simulations were performed using the M1 closure scheme, resolved with twelve frequency (energy) bins logarithmically spaced from $5 \times 10^{-3}$ to $5 \times 10^3$ keV. We apply a general relativistic Monte Carlo transport code to post-process the simulation data with greater fidelity in frequency resolution and Compton scattering treatment. Despite the relatively few energy bins and Kompaneets approximation to Compton scattering utilized in the M1 method, we find generally good agreement between the methods. Both produce prominent thermal profiles with peaks around 2 - 2.5 keV, where agreement is particularly strong and representative of the soft state. Both also find weaker (lower luminosity) thermally sourced emission extending out to 100 keV due to the hotter innermost regions of the disks. Inverse Compton scattering becomes increasingly effective at hardening spectral outputs with increasing black hole spin, and becomes the dominant mechanism for photons that escape with energies between 10 to several hundred keV. At very high rates of spin the radiation flux in this upscattered component becomes comparable to the thermal flux, a phenomenon typically associated with intermediate states. Beyond $10^4$ keV, we observe faint, free-free emission from hot, optically thin coronal regions developing near the horizon, common to both spinning and nonspinning black holes., Comment: 17 pages, 7 figures, ApJ

Effective token compression remains a critical challenge for scaling models to handle increasingly complex and diverse datasets. A novel mechanism based on contextual reinforcement is introduced, dynamically adjusting token importance through interdependencies and semantic relevance. This approach enables substantial reductions in token usage while preserving the quality and coherence of information representation. Incorporating graph-based algorithms and adaptive weighting, the method captures subtle contextual relationships across textual and multimodal data, ensuring robust alignment and performance in downstream tasks. Evaluations across varied domains reveal significant improvements in accuracy and semantic retention, particularly for tasks requiring detailed cross-modal interactions. Memory usage analyses demonstrate improved computational efficiency, with minimal overhead despite the additional reinforcement processes. Performance gains are further validated through error distribution analyses, showing reduced semantic loss and syntactic inconsistencies compared to baseline models. The modular architecture ensures compatibility with a wide range of open-source frameworks, facilitating scalable implementation for real-world applications. These findings highlight the potential of contextual reinforcement in redefining token management strategies and advancing large-scale model design.

Most of the world's languages and dialects are low-resource, and lack support in mainstream machine translation (MT) models. However, many of them have a closely-related high-resource language (HRL) neighbor, and differ in linguistically regular ways from it. This underscores the importance of model robustness to dialectical variation and cross-lingual generalization to the HRL dialect continuum. We present DialUp, consisting of a training-time technique for adapting a pretrained model to dialectical data (M->D), and an inference-time intervention adapting dialectical data to the model expertise (D->M). M->D induces model robustness to potentially unseen and unknown dialects by exposure to synthetic data exemplifying linguistic mechanisms of dialectical variation, whereas D->M treats dialectical divergence for known target dialects. These methods show considerable performance gains for several dialects from four language families, and modest gains for two other language families. We also conduct feature and error analyses, which show that language varieties with low baseline MT performance are more likely to benefit from these approaches., Comment: 9 pages, 46 incl. appendix

Historical linguists have long written "programs" that convert reconstructed words in an ancestor language into their attested descendants via ordered string rewrite functions (called sound laws) However, writing these programs is time-consuming, motivating the development of automated Sound Law Induction (SLI) which we formulate as Programming by Examples (PBE) with Large Language Models (LLMs) in this paper. While LLMs have been effective for code generation, recent work has shown that PBE is challenging but improvable by fine-tuning, especially with training data drawn from the same distribution as evaluation data. In this paper, we create a conceptual framework of what constitutes a "similar distribution" for SLI and propose four kinds of synthetic data generation methods with varying amounts of inductive bias to investigate what leads to the best performance. Based on the results we create a SOTA open-source model for SLI as PBE (+6% pass rate with a third of the parameters of the second-best LLM) and also highlight exciting future directions for PBE research.

A periodically loaded waveguide composed of periodic discrete nonlinear gain and radiating elements supports a stable oscillation regime related to the presence of an exceptional point of degeneracy (EPD). After reaching saturation, the EPD in the system establishes the oscillation frequency. We demonstrate a synchronization regime at a stable oscillation frequency, resulting in uniform saturated gain across the array and uniform radiating power. Unlike conventional one-dimensional cavity resonances, the oscillation frequency is independent of the array length. Our investigations further show that when small-signal gain is non-uniformly distributed across the array, the saturated gain results in having a uniform distribution at a gain value that generates an EPD. Experimental validation using the measured board confirmed that the system saturates at an EPD, with a measured spectrum exhibiting very low phase noise. This low noise allows for operation at a clean oscillation frequency. Additionally, the measured uniform power across the array corresponds to the simulation results. The proposed scheme can pave the way for a new generation of high-power radiating arrays with distributed active elements.

Vision language models have achieved impressive results across various fields. However, adoption in remote sensing remains limited, largely due to the scarcity of paired image-text data. To bridge this gap, synthetic caption generation has gained interest, traditionally relying on rule-based methods that use metadata or bounding boxes. While these approaches provide some description, they often lack the depth needed to capture complex wide-area scenes. Large language models (LLMs) offer a promising alternative for generating more descriptive captions, yet they can produce generic outputs and are prone to hallucination. In this paper, we propose a new method to enhance vision-language datasets for remote sensing by integrating maps as external data sources, enabling the generation of detailed, context-rich captions. Additionally, we present methods to measure and mitigate hallucinations in LLM-generated text. We introduce fMoW-mm, a multimodal dataset incorporating satellite imagery, maps, metadata, and text annotations. We demonstrate its effectiveness for automatic target recognition in few-shot settings, achieving superior performance compared to other vision-language remote sensing datasets.

Benchmarks are important tools for tracking the rapid advancements in large language model (LLM) capabilities. However, benchmarks are not keeping pace in difficulty: LLMs now achieve over 90\% accuracy on popular benchmarks like MMLU, limiting informed measurement of state-of-the-art LLM capabilities. In response, we introduce Humanity's Last Exam (HLE), a multi-modal benchmark at the frontier of human knowledge, designed to be the final closed-ended academic benchmark of its kind with broad subject coverage. HLE consists of 3,000 questions across dozens of subjects, including mathematics, humanities, and the natural sciences. HLE is developed globally by subject-matter experts and consists of multiple-choice and short-answer questions suitable for automated grading. Each question has a known solution that is unambiguous and easily verifiable, but cannot be quickly answered via internet retrieval. State-of-the-art LLMs demonstrate low accuracy and calibration on HLE, highlighting a significant gap between current LLM capabilities and the expert human frontier on closed-ended academic questions. To inform research and policymaking upon a clear understanding of model capabilities, we publicly release HLE at https://lastexam.ai., Comment: 26 pages, 6 figures

There are currently no psychometrically valid tools to measure the perceived danger of robots. To fill this gap, we provided a definition of perceived danger and developed and validated a 12-item bifactor scale through four studies. An exploratory factor analysis revealed four subdimensions of perceived danger: affective states, physical vulnerability, ominousness, and cognitive readiness. A confirmatory factor analysis confirmed the bifactor model. We then compared the perceived danger scale to the Godspeed perceived safety scale and found that the perceived danger scale is a better predictor of empirical data. We also validated the scale in an in-person setting and found that the perceived danger scale is sensitive to robot speed manipulations, consistent with previous empirical findings. Results across experiments suggest that the perceived danger scale is reliable, valid, and an adequate predictor of both perceived safety and perceived danger in human-robot interaction contexts., Comment: 9 pages, 2 figures, to be published in the Proceedings of the 2025 ACM/IEEE International Conference on Human-Robot Interaction (HRI)

We observed the Array Canary, a novel JavaScript anti-analysis technique currently exploited in-the-wild by the Phishing-as-a-Service framework Darcula. The Array Canary appears to be an advanced form of the array shuffling techniques employed by the Emotet JavaScript downloader. In practice, a series of Array Canaries are set within a string array and if modified will cause the program to endlessly loop. In this paper, we demonstrate how an Array Canary works and discuss Autonomous Function Call Resolution (AFCR), which is a method we created to bypass Array Canaries. We also introduce Arphsy, a proof-of-concept for AFCR designed to guide Large Language Models and security researchers in the deobfuscation of "canaried" JavaScript code. We accomplish this by (i) Finding and extracting all Immediately Invoked Function Expressions from a canaried file, (ii) parsing the file's Abstract Syntax Tree for any function that does not implement imported function calls, (iii) identifying the most reassigned variable and its corresponding function body, (iv) calculating the length of the largest string array and uses it to determine the offset values within the canaried file, (v) aggregating all the previously identified functions into a single file, and (vi) appending driver code into the verified file and using it to deobfuscate the canaried file.

Many applications, including medical diagnostics, sonar and navigation rely on the detection of acoustic waves. Photonic hydrophones demonstrate comparable sensitivity to piezoelectric-based hydrophones, but with significantly reduced size, weight and power requirements. In this paper we demonstrate a micron-sized free-standing silicon photonic hydrophone. We demonstrate sensitivity on the order of $\sim$mPa/$\sqrt{\textrm{Hz}}$ from 10-200 kHz, with a minimum detectable pressure of 145 $\mu$Pa/$\sqrt{\textrm{Hz}}$ at 22 kHz. We also deployed our hydrophone in a wave flume to evaluate its suitability for underwater measurement and communication. Our hydrophone matches the sensitivity of commercial hydrophones, but is many orders of magnitude smaller in volume, which could enable high spatial resolution imaging of micron-sized acoustic features (i.e., living cell vibrations). Our hydrophone could also be used in underwater communication and imaging applications., Comment: 11 pages, 5 figures

Surgical learning curves are graphical tools used to evaluate a trainee's progress in the early stages of their career and determine whether they have achieved proficiency after completing a specified number of surgeries. Cumulative sum (CUSUM) techniques are commonly used to assess learning curves due to their simplicity, but they face criticism for relying on fixed performance thresholds and lacking interpretability. This paper introduces a risk-adjusted surgical learning curve assessment (SLCA) method that focuses on estimation rather than hypothesis testing, as seen in CUSUM methods. The method is designed to accommodate right-skewed outcomes, such as surgery durations, characterized by the Weibull distribution. To evaluate the learning process, the SLCA approach estimates comparative probability metrics that assess the likelihood of a clinically important difference between the trainee's performance and a standard. Expecting improvement over time, we use weighted estimating equations to give greater weight to recent outcomes. Compared to CUSUM methods, SLCA offers enhanced interpretability, avoids reliance on externally defined performance levels, and emphasizes assessing clinical equivalence or noninferiority. We demonstrate the method's effectiveness through a colorectal surgery dataset case study and a numerical study.

Building upon the success of optimal transport metrics defined for single collinear jets, we develop a multi-scale framework that models entire collider events as distributions on the manifold of their constituent jets, which are themselves distributions on the ground space of the calorimeter. This hierarchical structure of optimal transport effectively captures relevant physics at different scales. We demonstrate the versatility of our method in two event classification tasks, which respectively emphasize intra-jet substructure and inter-jet spatial correlations. Our results highlight the relevance of a nested structure of manifolds in the treatment of full collider events, broadening the applicability of optimal transport methods in collider analyses., Comment: 12 pages, 9 figures

We present MOFA, an open-source generative AI (GenAI) plus simulation workflow for high-throughput generation of metal-organic frameworks (MOFs) on large-scale high-performance computing (HPC) systems. MOFA addresses key challenges in integrating GPU-accelerated computing for GPU-intensive GenAI tasks, including distributed training and inference, alongside CPU- and GPU-optimized tasks for screening and filtering AI-generated MOFs using molecular dynamics, density functional theory, and Monte Carlo simulations. These heterogeneous tasks are unified within an online learning framework that optimizes the utilization of available CPU and GPU resources across HPC systems. Performance metrics from a 450-node (14,400 AMD Zen 3 CPUs + 1800 NVIDIA A100 GPUs) supercomputer run demonstrate that MOFA achieves high-throughput generation of novel MOF structures, with CO$_2$ adsorption capacities ranking among the top 10 in the hypothetical MOF (hMOF) dataset. Furthermore, the production of high-quality MOFs exhibits a linear relationship with the number of nodes utilized. The modular architecture of MOFA will facilitate its integration into other scientific applications that dynamically combine GenAI with large-scale simulations., Comment: 13 pages, 10 figures

In many real-world applications, agents must make sequential decisions in environments where conditions are subject to change due to various exogenous factors. These non-stationary environments pose significant challenges to traditional decision-making models, which typically assume stationary dynamics. Non-stationary Markov decision processes (NS-MDPs) offer a framework to model and solve decision problems under such changing conditions. However, the lack of standardized benchmarks and simulation tools has hindered systematic evaluation and advance in this field. We present NS-Gym, the first simulation toolkit designed explicitly for NS-MDPs, integrated within the popular Gymnasium framework. In NS-Gym, we segregate the evolution of the environmental parameters that characterize non-stationarity from the agent's decision-making module, allowing for modular and flexible adaptations to dynamic environments. We review prior work in this domain and present a toolkit encapsulating key problem characteristics and types in NS-MDPs. This toolkit is the first effort to develop a set of standardized interfaces and benchmark problems to enable consistent and reproducible evaluation of algorithms under non-stationary conditions. We also benchmark six algorithmic approaches from prior work on NS-MDPs using NS-Gym. Our vision is that NS-Gym will enable researchers to assess the adaptability and robustness of their decision-making algorithms to non-stationary conditions., Comment: 23 pages, 17 figures

The use of structured light to drive highly nonlinear processes in matter not only enables imprinting spatially-resolved properties onto short-wavelength radiation, but also opens alternative avenues for exploring the dynamics of nonlinear laser-matter interactions. In this work, we experimentally and theoretically explore the unique properties of driving high-order harmonic generation (HHG) with Hermite-Gaussian beams. HHG driven by Laguerre-Gauss modes results in harmonics that inherit the azimuthal Laguerre-Gauss modal structure, with their topological charge scaling according to orbital angular momentum conservation. In contrast, when HHG is driven by Hermite-Gauss beams, the harmonic modes do not show a direct correspondence to the driving modal profile. Our experimental measurements using HG$_{01}$ and HG$_{11}$ modes, which are in excellent agreement with our numerical simulations, show that the lobes of the Hermite-Gauss driving beams effectively produce a set of separate phase-locked harmonic beamlets which can interfere downstream. This self-interference, which can be adjusted through the relative position between the gas target and the driving beam focus, can be exploited for precision extreme-ultraviolet interferometry. We demonstrate a simple application to calibrate the dispersion of an extreme-ultraviolet diffraction grating. In addition, we show through simulations that the array of harmonic beamlets can be used as an illumination source for single-shot extreme-ultraviolet ptychography., Comment: 14 pages, 8 figures

The quantum anomalous Hall effect (QAHE) in magnetic topological insulators offers great potential to revolutionize quantum electrical metrology by establishing primary resistance standards operating at zero external magnetic field and realizing a universal "quantum electrical metrology toolbox" that can perform quantum resistance, voltage and current metrology in a single instrument. To realize such promise, significant progress is still required to address materials and metrological challenges -- among which, one main challenge is to make the bulk of the topological insulator sufficiently insulating to improve the robustness of resistance quantization. In this Perspective, we present an overview of the QAHE; discuss the aspects of topological material growth and characterization; and present a path towards an QAHE resistance standard realized in magnetically doped (Bi,Sb)$_2$Te$_3$ systems. We also present guidelines and methodologies for QAHE resistance metrology, its main limitations and challenges as well as modern strategies to overcome them., Comment: 15 pages, 4 figures

Music source separation demixes a piece of music into its individual sound sources (vocals, percussion, melodic instruments, etc.), a task with no simple mathematical solution. It requires deep learning methods involving training on large datasets of isolated music stems. The most commonly available datasets are made from commercial Western music, limiting the models' applications to non-Western genres like Carnatic music. Carnatic music is a live tradition, with the available multi-track recordings containing overlapping sounds and bleeds between the sources. This poses a challenge to commercially available source separation models like Spleeter and Hybrid Demucs. In this work, we introduce 'Sanidha', the first open-source novel dataset for Carnatic music, offering studio-quality, multi-track recordings with minimal to no overlap or bleed. Along with the audio files, we provide high-definition videos of the artists' performances. Additionally, we fine-tuned Spleeter, one of the most commonly used source separation models, on our dataset and observed improved SDR performance compared to fine-tuning on a pre-existing Carnatic multi-track dataset. The outputs of the fine-tuned model with 'Sanidha' are evaluated through a listening study., Comment: Accepted to the 25th International Society for Music Information Retrieval Conference (ISMIR 2024)

Recent research using Reinforcement Learning (RL) to learn autonomous control for spacecraft operations has shown great success. However, a recent study showed their performance could be improved by changing the action space, i.e. control outputs, used in the learning environment. This has opened the door for finding more improvements through further changes to the environment. The work in this paper focuses on how changes to the environment's observation space can impact the training and performance of RL agents learning the spacecraft inspection task. The studies are split into two groups. The first looks at the impact of sensors that were designed to help agents learn the task. The second looks at the impact of reference frames, reorienting the agent to see the world from a different perspective. The results show the sensors are not necessary, but most of them help agents learn more optimal behavior, and that the reference frame does not have a large impact, but is best kept consistent., Comment: 18 pages, 10 figures, 3 tables

The Safe Trusted Autonomy for Responsible Space (STARS) program aims to advance autonomy technologies for space by leveraging machine learning technologies while mitigating barriers to trust, such as uncertainty, opaqueness, brittleness, and inflexibility. This paper presents the achievements and lessons learned from the STARS program in integrating reinforcement learning-based multi-satellite control, run time assurance approaches, and flexible human-autonomy teaming interfaces, into a new integrated testing environment for collaborative autonomous satellite systems. The primary results describe analysis of the reinforcement learning multi-satellite control and run time assurance algorithms. These algorithms are integrated into a prototype human-autonomy interface using best practices from human-autonomy trust literature, however detailed analysis of the effectiveness is left to future work. References are provided with additional detailed results of individual experiments.

Given a stochastic process with inputs and outputs, how might its behaviour be related to pursuit of a goal? We model this using 'transducers', objects that capture only the external behaviour of a system and not its internal state. A companion paper summarises our results for cognitive scientists; the current paper gives formal definitions and proofs. To formalise the concept of a system that behaves as if it were pursuing a goal, we consider what happens when a transducer (a 'policy') is coupled to another transducer that comes equipped with a success condition (a 'teleo-environment'). An optimal policy is identified with a transducer that behaves as if it were perfectly rational in the pursuit of a goal; our framework also allows us to model constrained rationality. Optimal policies obey a version of Bellman's principle: a policy that's optimal in one time step will again be optimal in the next time step, but with respect to a different teleo-environment (obtained from the original one by a modified version of Bayesian filtering). This property sometimes also applies to the bounded-rational case; we give a sufficient condition. A policy is deterministic if and only if there exists a teleo-environment for which it is uniquely optimal among the set of all policies; we relate this to classical representation theorems from decision theory. This result need not hold in the bounded-rational case; we give an example related to the absent-minded driver problem. The formalism is defined using coinduction, following the style proposed by Czajka., Comment: This is the companion paper to "Formalising the intentional stance 1: attributing goals and beliefs to stochastic processes" (uploaded as version 2 of arXiv:2405.16490). The other paper is an overview aimed at cognitive scientists while this paper gives full mathematical details. 50 pages, no figures

Hair care robots can help address labor shortages in elderly care while enabling those with limited mobility to maintain their hair-related identity. We present MOE-Hair, a soft robot system that performs three hair-care tasks: head patting, finger combing, and hair grasping. The system features a tendon-driven soft robot end-effector (MOE) with a wrist-mounted RGBD camera, leveraging both mechanical compliance for safety and visual force sensing through deformation. In testing with a force-sensorized mannequin head, MOE achieved comparable hair-grasping effectiveness while applying significantly less force than rigid grippers. Our novel force estimation method combines visual deformation data and tendon tensions from actuators to infer applied forces, reducing sensing errors by up to 60.1% and 20.3% compared to actuator current load-only and depth image-only baselines, respectively. A user study with 12 participants demonstrated statistically significant preferences for MOE-Hair over a baseline system in terms of comfort, effectiveness, and appropriate force application. These results demonstrate the unique advantages of soft robots in contact-rich hair-care tasks, while highlighting the importance of precise force control despite the inherent compliance of the system.

This paper presents a novel deep learning-based approach named RealDiffFusionNet incorporating Neural Controlled Differential Equations (Neural CDE) - time series models that are robust in handling irregularly sampled data - and multi-head attention to align relevant multimodal context (image data, time invariant data, etc.) at each time point. Long short-term memory (LSTM) models were also used as a baseline. Two different datasets were used: a data from the Open-Source Imaging Consortium (OSIC) containing structured time series data of demographics and lung function with a baseline CT scan of the lungs and the second from the Alzheimer's Disease Neuroimaging Initiative (ADNI) containing a series of MRI scans along with demographics, physical examinations, and cognitive assessment data. An ablation study was performed to understand the role of CDEs, multimodal data, attention fusion, and interpolation strategies on model performance. When the baseline models were evaluated, the use of multimodal data resulted in an improvement in Neural CDE performance, with a lower test RMSE. Additionally, the performance of multimodal Neural CDE was also superior to multimodal LSTM. In the attention-based architectures, fusion through concatenation and rectilinear interpolation were found to improve model performance. The performance of the proposed RealDiffFusionNet was found to be superior (0.2570) to all models. For the ADNI dataset, between the Neural-CDE and LSTM models trained only on the structured data, the test RMSE were comparable (0.471 for LSTM vs. 0.4581 Neural-CDE). Furthermore, the addition of image features from patients' MRI series resulted in an improvement in performance, with a lower test RMSE (0.4372 with multimodal vs 0.4581 with structured data). RealDiffFusionNet has shown promise in utilizing CDEs and multimodal data to accurately predict disease progression.

People have a variety of preferences for how robots behave. To understand and reason about these preferences, robots aim to learn a reward function that describes how aligned robot behaviors are with a user's preferences. Good representations of a robot's behavior can significantly reduce the time and effort required for a user to teach the robot their preferences. Specifying these representations -- what "features" of the robot's behavior matter to users -- remains a difficult problem; Features learned from raw data lack semantic meaning and features learned from user data require users to engage in tedious labeling processes. Our key insight is that users tasked with customizing a robot are intrinsically motivated to produce labels through exploratory search; they explore behaviors that they find interesting and ignore behaviors that are irrelevant. To harness this novel data source of exploratory actions, we propose contrastive learning from exploratory actions (CLEA) to learn trajectory features that are aligned with features that users care about. We learned CLEA features from exploratory actions users performed in an open-ended signal design activity (N=25) with a Kuri robot, and evaluated CLEA features through a second user study with a different set of users (N=42). CLEA features outperformed self-supervised features when eliciting user preferences over four metrics: completeness, simplicity, minimality, and explainability., Comment: Accepted to HRI 2025

We present new observations of the cosmic ultraviolet background (CUVB) at high Galactic latitudes ($|b| > 40^{\circ}$), made using the Alice UV spectrograph on board the New Horizons spacecraft. These observations were taken at about 57 AU from the Sun, outside much of the foreground emission affecting previous missions, and allowed a new determination of the spectrum of the CUVB between 912 -- 1100~\AA\ and 1400 -- 1800~\AA. We found a linear correlation between the CUVB and the Planck E(B~-~V) with offsets at zero-reddening of $221 \pm 11$ photon units at 1000~\AA\ and $264 \pm 24$ \photu\ at 1500~\AA\ ($4.4 \pm 0.2$ nW m$^{-2}$ sr$^{-1}$ at 1000~\AA\ and $5.3 \pm 0.5$ nW m$^{-2}$ sr$^{-1}$ at 1500~\AA). The former is the first firm detection of the offset in the range 912 -- 1100 \AA\ while the latter result confirms previous results from \galex, showing that there is little emission from the Solar System from 1400 -- 1800 \AA. About half of the offset may be explained by known sources (the integrated light of unresolved galaxies, unresolved stars, emission from ionized gas, and two-photon emission from warm hydrogen in the halo) with the source of the remaining emission as yet unidentified. There is no detectable emission below the Lyman limit with an upper limit of $3.2 \pm 3.0$ photon units., Comment: Accepted in AJ. 19 pages, 16 figures, 8 tables

This note relates to the author's construction of the Symplectic $(A_\infty,2)$-Category, $\mathsf{Symp}$. Here we explain two ways of encoding the information in $\mathsf{Symp}$, one topological, one algebraic. The topological encoding is as an $(A_\infty,2)$-flow category, which we define here. The algebraic encoding is as a linear $(A_\infty,2)$-category, which we extract from the topological encoding. In upcoming work, the author and Wehrheim plan to use the adiabatic Fredholm theory recently developed by Bottman-Wehrheim to construct $\mathsf{Symp}$ as an $(A_\infty,2)$-flow category. The definition of linear $(A_\infty,2)$-category that we give in this note is different than the one proposed by Bottman-Carmeli. The recursive structure of the 2-associahedra identifies faces with fiber products of 2-associahedra over associahedra, and these fiber products led Bottman-Carmeli to associate operations to singular chains on 2-associahedra. The innovation in our new definition of linear $(A_\infty,2)$-category is to extend the family of 2-associahedra to include all fiber products of 2-associahedra over associahedra. This allows us to associate operations to cellular chains, which in particular enables us to produce a definition that involves only one operation in each arity, governed by a collection of $(A_\infty,2)$-equations., Comment: 11 pages, 1 figure

Recent reasoning methods (e.g., chain-of-thought, entailment reasoning) help users understand how language models (LMs) answer a single question, but they do little to reveal the LM's overall understanding, or "theory," about the question's topic, making it still hard to trust the model. Our goal is to materialize such theories - here called microtheories (a linguistic analog of logical microtheories) - as a set of sentences encapsulating an LM's core knowledge about a topic. These statements systematically work together to entail answers to a set of questions to both engender trust and improve performance. Our approach is to first populate a knowledge store with (model-generated) sentences that entail answers to training questions and then distill those down to a core microtheory that is concise, general, and non-redundant. We show that, when added to a general corpus (e.g., Wikipedia), microtheories can supply critical, topical information not necessarily present in the corpus, improving both a model's ability to ground its answers to verifiable knowledge (i.e., show how answers are systematically entailed by documents in the corpus, fully grounding up to +8% more answers), and the accuracy of those grounded answers (up to +8% absolute). We also show that, in a human evaluation in the medical domain, our distilled microtheories contain a significantly higher concentration of topically critical facts than the non-distilled knowledge store. Finally, we show we can quantify the coverage of a microtheory for a topic (characterized by a dataset) using a notion of $p$-relevance. Together, these suggest that microtheories are an efficient distillation of an LM's topic-relevant knowledge, that they can usefully augment existing corpora, and can provide both performance gains and an interpretable, verifiable window into the model's knowledge of a topic.

We prove that the uniform doubling property holds for every Lie group which can be written as a quotient group of $\operatorname{SU}(2) \times \mathbb{R}^n$ for some $n$. In particular, this class includes the four-dimensional unitary group $\operatorname{U}(2)$. As this class contain non-compact as well as compact Lie groups, we discuss a number of analytic and spectral consequences for the corresponding heat kernels., Comment: 35 pages

Exchange-only (EO) spin qubits in quantum dots offer an expansive design landscape for architecting scalable device layouts. The study of two-EO-qubit operations, which involve six electrons in six quantum dots, has so far been limited to a small number of the possible configurations, and previous works lack analyses of design considerations and implications for quantum error correction. Using a simple and fast optimization method, we generate complete pulse sequences for CX, CZ, iSWAP, leakage-controlled CX, and leakage-controlled CZ two-qubit gates on 450 unique planar six-dot topologies and analyze differences in sequence length (up to 43% reduction) across topology classes. In addition, we show that relaxing constraints on post-operation spin locations can yield further reductions in sequence length; conversely, constraining these locations in a particular way generates a CXSWAP operation with minimal additional cost over a standard CX. We integrate this pulse library into the Intel quantum stack and experimentally verify pulse sequences on a Tunnel Falls chip for different operations in a linear-connectivity device to confirm that they work as expected. Finally, we explore architectural implications of these results for quantum error correction. Our work guides hardware and software design choices for future implementations of scalable quantum dot architectures., Comment: 15 pages, 12 figures

We examine nonrenormalizable Lorentz- and CPT-violating effective operators applied to the quark sector of the Standard Model. Using Drell-Yan events collected by the ATLAS and CMS Collaborations, several constraints are extracted from time-independent modifications of the cross section on the $Z$-boson pole. The sensitivity to time-dependent modifications are also estimated by simulating a sidereal-time analysis. Our results suggest a dedicated search can improve on constraints from deep inelastic scattering by up to three orders in magnitude., Comment: 7 pages

The Ni-Au-Ga interdiffusion mechanisms taking place during rapid thermal annealing (RTA) under oxygen atmosphere of a Ni-Au/p-GaN contact are investigated by high-resolution transmission electron microscopy (HR-TEM) coupled to energy dispersive X-ray spectroscopy (EDX). It is shown that oxygen-assisted, Ni diffusion to the top surface of the metallic contact through the formation of a nickel oxide (NiOx) is accompanied by Au diffusion down to the GaN surface, and by Ga out-diffusion through the GaN/metal interface. Electrical characterizations of the contact by Transmission Line Method (TLM) show that an ohmic contact is obtained as soon as a thin, Au-Ga interfacial layer is formed, even after complete diffusion of Ni or NiOx to the top surface of the contact. Our results clarify that the presence of Ni or NiOx at the interface is not the main origin of the ohmic-like behavior in such contacts. Auto-cleaning of the interface during the interdiffusion process may play a role, but TEM-EDX analysis evidences that the creation of Ga vacancies associated to the formation of a Ga-Au interfacial layer is crucial for reducing the Schottky barrier height, and maximizing the amount of current flowing through the contact.

Given an edge-colored graph, the goal of the proportional fair matching problem is to find a maximum weight matching while ensuring proportional representation (with respect to the number of edges) of each color. The colors may correspond to demographic groups or other protected traits where we seek to ensure roughly equal representation from each group. It is known that, assuming ETH, it is impossible to approximate the problem with $\ell$ colors in time $2^{o(\ell)} n^{\mathcal{O}(1)}$ (i.e., subexponential in $\ell$) even on \emph{unweighted path graphs}. Further, even determining the existence of a non-empty matching satisfying proportionality is NP-Hard. To overcome this hardness, we relax the stringent proportional fairness constraints to a probabilistic notion. We introduce a notion we call $\delta$-\textsc{ProbablyAlmostFair}, where we ensure proportionality up to a factor of at most $(1 \pm \delta)$ for some small $\delta >0$ with high probability. The violation $\delta$ can be brought arbitrarily close to $0$ for some \emph{good} instances with large values of matching size. We propose and analyze simple and fast algorithms for bipartite graphs that achieve constant-factor approximation guarantees, and return a $\delta$-\textsc{ProbablyAlmostFair} matching.

As the number of spacecraft in orbit continues to increase, it is becoming more challenging for human operators to manage each mission. As a result, autonomous control methods are needed to reduce this burden on operators. One method of autonomous control is Reinforcement Learning (RL), which has proven to have great success across a variety of complex tasks. For missions with multiple controlled spacecraft, or agents, it is critical for the agents to communicate and have knowledge of each other, where this information is typically given to the Neural Network Controller (NNC) as an input observation. As the number of spacecraft used for the mission increases or decreases, rather than modifying the size of the observation, this paper develops a scalable observation space that uses a constant observation size to give information on all of the other agents. This approach is similar to a lidar sensor, where determines ranges of other objects in the environment. This observation space is applied to a spacecraft inspection task, where RL is used to train multiple deputy spacecraft to cooperate and inspect a passive chief spacecraft. It is expected that the scalable observation space will allow the agents to learn to complete the task more efficiently compared to a baseline solution where no information is communicated between agents.

Phase space engineering by RF waves plays important roles in both thermal D-T fusion and non-thermal advanced fuel fusion. But not all phase space manipulation is allowed, certain fundamental limits exist. In addition to Liouville's theorem, which requires the manipulation to be volume-preserving, Gromov's non-squeezing theorem imposes another constraint. The Gardner ground state is defined as the ground state accessible by smooth volume-preserving maps. However, the extra Gromov constraint should produce a higher-energy ground state. An example of a Gardner ground state forbidden by Gromov's non-squeezing theorem is given. The challenge question is: What is the Gromov ground state, i.e., the lowest energy state accessible by smooth symplectic maps? This is a difficult problem. As a simplification, we conjecture that the linear Gromov ground state problem is solvable., Comment: 17 pages, 1 figure

We present a scalable, robust approach to creating quantum programs of arbitrary size and complexity. The approach is based on the true abstraction of the problem. The quantum program is expressed in terms of a high-level model together with constraints and objectives on the final program. Advanced synthesis algorithms transform the model into a low-level quantum program that meets the user's specification and is directed at a stipulated hardware. This separation of description from implementation is essential for scale. The technology adapts electronic design automation methods to quantum computing, finding feasible implementations in a virtually unlimited functional space. The results show clear superiority over the compilation and transpilation methods used today. We expect that this technological approach will take over and prevail as quantum software become more demanding, complex, and essential.

Dynamical phase transitions (DPTs) characterize critical changes in system behavior occurring at finite times, providing a lens to study nonequilibrium phenomena beyond conventional equilibrium physics. While extensively studied in quantum systems, DPTs have remained largely unexplored in classical settings. Recent experiments on complex systems, from social networks to financial markets, have revealed abrupt dynamical changes analogous to quantum DPTs, motivating the search for a theoretical understanding. Here, we present a minimal model for nonequilibrium networks, demonstrating that nonlinear interactions among network edges naturally give rise to DPTs. Specifically, we show that network degree diverges at a finite critical time, following a universal hyperbolic scaling, consistent with empirical observations. Our analytical results predict that key network properties, including degree distributions and clustering coefficients, exhibit critical scaling as criticality approaches. These findings establish a theoretical foundation for understanding emergent nonequilibrium criticality across diverse complex systems.

Physical data are representations of the fundamental laws governing the Universe, hiding complex compositional structures often well captured by hierarchical graphs. Hyperbolic spaces are endowed with a non-Euclidean geometry that naturally embeds those structures. To leverage the benefits of non-Euclidean geometries in representing natural data we develop machine learning on $\mathcal P \mathcal M$ spaces, Cartesian products of constant curvature Riemannian manifolds. As a use case we consider the classification of "jets", sprays of hadrons and other subatomic particles produced by the hadronization of quarks and gluons in collider experiments. We compare the performance of $\mathcal P \mathcal M$-MLP and $\mathcal P \mathcal M$-Transformer models across several possible representations. Our experiments show that $\mathcal P \mathcal M$ representations generally perform equal or better to fully Euclidean models of similar size, with the most significant gains found for highly hierarchical jets and small models. We discover significant correlation between the degree of hierarchical structure at a per-jet level and classification performance with the $\mathcal P \mathcal M$-Transformer in top tagging benchmarks. This is a promising result highlighting a potential direction for further improving machine learning model performance through tailoring geometric representation at a per-sample level in hierarchical datasets. These results reinforce the view of geometric representation as a key parameter in maximizing both performance and efficiency of machine learning on natural data.

In a magnetic mirror fusion reactor, capturing the energy of fusion-produced alpha particles is essential to sustaining the reaction. However, since alpha particles are born at energies much higher than the confining potential, a substantial fraction are lost due to pitch-angle scattering before they can transfer their energy to the plasma via drag. The energy of prematurely lost alpha particles can still be captured through direct conversion, but designing an effective mechanism requires a description of the energies and times at which they become deconfined. Here we present the first-ever analytical solutions for the loss velocity, energy, and time distributions of alpha particles in a magnetic mirror. After obtaining the Fokker-Planck model of the Landau collision operator, we asymptotically solve for the orthogonal eigenfunctions of the Legendre operator with homogeneous boundary conditions to reveal an easily interpretable closed-form solution. Our framework is extremely general and can in principle be applied to any high-energy species, subject to any applied potential, in the low collisionality, modest potential regime for any mirror ratio R > 1, making this work broadly applicable to existing mirror devices., Comment: 12 pages, 15 figures