Search

Since 2019, the world has irrevocably changed, and public education is not exempt. The pandemic and related political, social, economic, and technological developments have indelibly changed the K-12 landscape. Over the past five years, the Canopy Project has documented innovations in K-12 education, uncovering how schools are addressing systemic challenges and envisioning the future of learning. This report highlights the progress and persistence of these trailblazing schools and offers actionable recommendations for sustaining and spreading innovation. To improve student experiences and outcomes on a large scale and in schools beyond the Canopy project, policymakers, funders, and researchers must play critical roles in strengthening the conditions for innovation in schools. This report demonstrates that reimagining learning environments in sustainable, school-wide ways is achievable in any school, anywhere in the country. This transformation is already underway, and students are reaping the benefits. However, for school innovation to endure and expand, policymakers, funders, and researchers must support the conditions necessary. [This report was produced with Transcend.]

In the face of financial, political, and capacity constraints, leaders within the Baltimore City Public School System and Chicago Public Schools are making progress toward closing post-pandemic gaps in student learning. The executive leadership teams serving Baltimore City Public School System and Chicago Public Schools are using several strategies to help their systems build back after the pandemic including: (1) standardizing key elements of instructional cores and adding instructional staff across their systems, while maintaining some school-level flexibility; (2) supporting existing staff to become teachers and "building the bench" of new principals, while also promoting school leaders and administrators from within; (3) spending federal pandemic-relief money Elementary and Secondary School Emergency Relief Fund (ESSER) and other supplements, while hoping to avoid cutting into funding for instruction; (4) insulating schools as much as possible by taking the political hits themselves and making concessions to demanding stakeholders; and (5) confronting continued enrollment declines, limited central office capacity, constrained teacher labor markets, changes in school boards, shifting local interests, and uncertain state-level financial support. The strong leadership modeled in both districts remains at the heart of moving forward, despite the fact they are straining against their capacities. While there may be nuanced differences, all leaders in maxed-out districts will have to search for new options and forge alliances with new partners within school boards, unions, and state-level governments. This project is part of the American School District Panel (ASDP), a research partnership between the RAND Corporation and the Center on Reinventing Public Education (CRPE) at Arizona State University's Mary Lou Fulton Teachers College.

Vision-language model (VLM) fine-tuning for application-specific visual grounding based on natural language instructions has become one of the most popular approaches for learning-enabled autonomous systems. However, such fine-tuning relies heavily on high-quality datasets to achieve successful performance in various downstream tasks. Additionally, VLMs often encounter limitations due to insufficient and imbalanced fine-tuning data. To address these issues, we propose a new generalizable framework to improve VLM fine-tuning by integrating it with a reinforcement learning (RL) agent. Our method utilizes the RL agent to manipulate objects within an indoor setting to create synthetic data for fine-tuning to address certain vulnerabilities of the VLM. Specifically, we use the performance of the VLM to provide feedback to the RL agent to generate informative data that efficiently fine-tune the VLM over the targeted task (e.g. spatial reasoning). The key contribution of this work is developing a framework where the RL agent serves as an informative data sampling tool and assists the VLM in order to enhance performance and address task-specific vulnerabilities. By targeting the data sampling process to address the weaknesses of the VLM, we can effectively train a more context-aware model. In addition, generating synthetic data allows us to have precise control over each scene and generate granular ground truth captions. Our results show that the proposed data generation approach improves the spatial reasoning performance of VLMs, which demonstrates the benefits of using RL-guided data generation in vision-language tasks., Comment: ICCPS 2025 accepted paper, 10 pages, 9 figures

Combining microstructural mechanical models with experimental data enhances our understanding of the mechanics of soft tissue, such as tendons. In previous work, a Bayesian framework was used to infer constitutive parameters from uniaxial stress-strain experiments on horse tendons, specifically the superficial digital flexor tendon (SDFT) and common digital extensor tendon (CDET), on a per-experiment basis. Here, we extend this analysis to investigate the natural variation of these parameters across a population of horses. Using a Bayesian mixed effects model, we infer population distributions of these parameters. Given that the chosen hyperelastic model does not account for tendon damage, careful data selection is necessary. Avoiding ad hoc methods, we introduce a hierarchical Bayesian data selection method. This two-stage approach selects data per experiment, and integrates data weightings into the Bayesian mixed effects model. Our results indicate that the CDET is stiffer than the SDFT, likely due to a higher collagen volume fraction. The modes of the parameter distributions yield estimates of the product of the collagen volume fraction and Young's modulus as 811.5 MPa for the SDFT and 1430.2 MPa for the CDET. This suggests that positional tendons have stiffer collagen fibrils and/or higher collagen volume density than energy-storing tendons., Comment: 16 pages, 9 figures

Climate change poses an urgent global threat, needing the rapid identification and deployment of innovative solutions. We hypothesise that many of these solutions already exist within scientific literature but remain underutilised. To address this gap, this study employs a curated dataset sourced from OpenAlex, a comprehensive repository of scientific papers. Utilising Large Language Models (LLMs), such as GPT4-o from OpenAI, we evaluate title-abstract pairs from scientific papers on seven dimensions, covering climate change mitigation potential, stage of technological development, and readiness for deployment. The outputs of the language models are then compared with human evaluations to assess their effectiveness in identifying promising yet overlooked climate innovations. Our findings suggest that these LLM-based models can effectively augment human expertise, uncovering climate solutions that are potentially impactful but with far greater speed, throughput and consistency. Here, we focused on UK-based solutions, but the workflow is region-agnostic. This work contributes to the discovery of neglected innovations in scientific literature and demonstrates the potential of AI in enhancing climate action strategies., Comment: 10 pages. Accepted in the LatinX in AI workshop at NeurIPS 2024

Recently we introduced a local photon approach for modelling the quantised electromagnetic field in position space. Using this approach, we define the momentum of light in this paper as in quantum mechanics as the generator for spatial translation. Afterwards, we analyse the momentum dynamics of photonic wave packets which transition from air into a denser dielectric medium. Our analysis shines new light onto the Abraham-Minkowski controversy which highlights the intricacies involved in the characterisation of the momentum of the electromagnetic field. Although our results align with Minkowski's theory and with the definition of the canonical momentum of light in quantum electrodynamics, there are also some crucial differences., Comment: 15 pages, 2 figures, new title, accepted version

The COVID-19 pandemic and recovery period is a unique opportunity to understand contemporary issues in high school reform. Evidence has clearly demonstrated lingering COVID impacts on adolescent students that have deepened pre-existing inequities and worsened teen mental health. There's a natural desire to regain normalcy after the pandemic. But it's essential, and urgent, that researchers examine why that "normalcy" failed in the past to support every student's needs. They must identify effective, even new, ways to level the playing field for today's students, and for future generations. Research has shown that many high school educators and administrators experimented with new approaches during the pandemic. Can the lessons learned in this period contribute to more lasting, transformative shifts? Researchers studied a group of New England public high schools during the pandemic recovery period to explore this question and hear directly from students and adults about their experiences. This report presents a set of recommendations for educational leaders and policy makers at the state and local levels -- along with caregivers, employers, higher education, and other community partners -- to help schools overcome barriers and think more broadly about how best to prepare graduates for a variety of paths to success after high school.

In this work, we explore the effect of neutrino nonstandard interactions (NSI) involving the charm quark at SND@LHC. Using an effective description of new physics in terms of four-fermion operators involving a charm quark, we constrain the Wilson coefficients of the effective interaction from two and three-body charmed meson decays. In our fit, we include charmed meson decays not only to pseudoscalar final states but also to vector final states and include decays to the $\eta$ and $\eta^\prime$ final states. We also consider constraints from charmed baryon decays. We then study the effect of new physics in neutrino scattering processes, involving charm production at SND@LHC, for various benchmark new physics couplings obtained from the low energy fits. Finally, we also study the effects of lepton universality violation (LUV) assuming that the new physics coupling is not lepton universal., Comment: 24 pages, 3 figures; final version published in Journal of High Energy Physics

Tendon-driven robotic catheters are capable of precise execution of minimally invasive cardiac procedures including ablations and imaging. These procedures require accurate mathematical models of not only the catheter and tendons but also their interactions with surrounding tissue and vasculature in order to control the robot path and interaction. This paper presents a mechanical model of a tendon-driven robotic catheter system based on Cosserat rods and integrated with a stable, implicit Euler scheme. We implement the Cosserat rod as a model for a simple catheter centerline and validate its physical accuracy against a large deformation analytical model and experimental data. The catheter model is then supplemented by adding a second Cosserat rod to model a single tendon, using penalty forces to define the constraints of the tendon-catheter system. All the model parameters are defined by the catheter properties established by the design. The combined model is validated against experimental data to confirm its physical accuracy. This model represents a new contribution to the field of robotic catheter modeling in which both the tendons and catheter are modeled by mechanical Cosserat rods and fully-validated against experimental data in the case of the single rod system., Comment: 24 pages, 23 figures

Recent advances in decentralized deep learning algorithms have demonstrated cutting-edge performance on various tasks with large pre-trained models. However, a pivotal prerequisite for achieving this level of competitiveness is the significant communication and computation overheads when updating these models, which prohibits the applications of them to real-world scenarios. To address this issue, drawing inspiration from advanced model merging techniques without requiring additional training, we introduce the Decentralized Iterative Merging-And-Training (DIMAT) paradigm--a novel decentralized deep learning framework. Within DIMAT, each agent is trained on their local data and periodically merged with their neighboring agents using advanced model merging techniques like activation matching until convergence is achieved. DIMAT provably converges with the best available rate for nonconvex functions with various first-order methods, while yielding tighter error bounds compared to the popular existing approaches. We conduct a comprehensive empirical analysis to validate DIMAT's superiority over baselines across diverse computer vision tasks sourced from multiple datasets. Empirical results validate our theoretical claims by showing that DIMAT attains faster and higher initial gain in accuracy with independent and identically distributed (IID) and non-IID data, incurring lower communication overhead. This DIMAT paradigm presents a new opportunity for the future decentralized learning, enhancing its adaptability to real-world with sparse and light-weight communication and computation., Comment: CVPR 2024 accepted paper, 22 pages, 12 figures

The temperature structure of a giant planet was traditionally thought to be an adiabat assuming convective mixing homogenizes entropy. The only in-situ measurement made by the Galileo Probe detected a near-adiabatic temperature structure within one of Jupiter's 5$\mu$m hot spots with small but definite local departures from adiabaticity. We analyze Juno's microwave observations near Jupiter's equator (0 ~ 5$^o$N) and find that the equatorial temperature structure is best characterized by a stable super-adiabatic temperature profile rather than an adiabatic one. Water is the only substance with sufficient abundance to alter the atmosphere's mean molecular weight and prevent dynamic instability if a super-adiabatic temperature gradient exists. Thus, from the super-adiabaticity, our results indicate a water concentration (or the oxygen to hydrogen ratio) of about 4.9 times solar with a possible range of 1.5 ~ 8.3 times solar in Jupiter's equatorial region., Comment: 46 pages, 11 figures, accepted by Icraus

This paper addresses the problem of data-driven modeling and verification of perception-based autonomous systems. We assume the perception model can be decomposed into a canonical model (obtained from first principles or a simulator) and a noise model that contains the measurement noise introduced by the real environment. We focus on two types of noise, benign and adversarial noise, and develop a data-driven model for each type using generative models and classifiers, respectively. We show that the trained models perform well according to a variety of evaluation metrics based on downstream tasks such as state estimation and control. Finally, we verify the safety of two systems with high-dimensional data-driven models, namely an image-based version of mountain car (a reinforcement learning benchmark) as well as the F1/10 car, which uses LiDAR measurements to navigate a racing track., Comment: 23 pages, 12 figures, and 3 tables. Submitted to: 6th Annual Learning for Dynamics & Control Conference

Nokomis Regional High School, which draws nearly 600 students from eight different towns in rural Maine, has spent 10 years transforming its instructional model to immerse students in meaningful learning experiences that relate to their interests and passions. Project-based learning, interdisciplinary courses, and career exploration units are designed to help students explore various options and feel secure about their postsecondary choices. As students near graduation in this rural community, some plan to enroll in college, while others envision entering the trades or starting a job. Staff at Nokomis want students to achieve success on their own terms, which means challenging the predominantly college-oriented attitude that persists in high school practice and policy. But many teachers and administrators are still grappling with how to ensure that students choosing different pathways--some involving college, some not (or not yet)--will have an equal shot at long-term success and stability relative to their peers. Some staff think the school should more assertively push students toward college and career options beyond the local rural community. This case wrestles with how high schools can expand students' postsecondary horizons while remaining responsive to their values and interests--especially if those values differ significantly from those of the adults who support them. [For the corresponding Case Studies in High School Redesign, see ED626304 and ED626305.]

Between 2015 and 2019, Holyoke Public Schools in Massachusetts cut its dropout rate nearly in half and increased its graduation rate by 10 percentage points, with notable gains among historically marginalized student groups. The district cites Opportunity Academy (OA), an alternative education program within the district's high school, as a key driver of these improvements. OA surrounds students with supportive relationships and provides multiple pathways to graduation, including a competency-based pathway in which students earn credits through flexible learning experiences inside and outside traditional classrooms. But for OA's design team, getting students to graduate isn't enough. Real success, the principal said, means students have "the ability to live a choice-filled postsecondary life." This case wrestles with what's needed to achieve this higher bar for students who often just want to get high school over with, but whose success beyond the K-12 system depends on life skills, supportive relationships, and concrete postsecondary plans in addition to a diploma. The case concludes with key questions for high school leaders and design teams in both traditional and alternative education. [For the corresponding Case Studies in High School Redesign, see ED626303 and ED626304.]