Data assimilation (DA) methods combine model predictions with observational data to improve state estimation in dynamical systems, inspiring their increasingly prominent role in geophysical and climate applications. Classical DA methods assume that the governing equations modeling the dynamics are known, which is unlikely for most real world applications. Machine learning (ML) provides a flexible alternative by learning surrogate models directly from data, but standard ML methods struggle in noisy and data-scarce environments, where meaningful extrapolation requires incorporating physical constraints. Recent advances in structure-preserving ML architectures, such as the development of the entropy-stable conservative flux form network (ESCFN), highlight the critical role of explicitly encoding physical structure (e.g., conservation laws and entropy conditions) into surrogate models, thereby improving learning stability and accuracy. Structural information has also been shown to improve DA performance. Gradient-based measures of spatial variability, in particular, can help refine ensemble updates in discontinuous systems. Motivated by both of these recent innovations, this investigation proposes a new non-intrusive, structure-preserving sequential data assimilation (NSSDA) framework that leverages structure at both the forecast and analysis stages. We use the ESCFN to construct a surrogate model to preserve physical laws during forecasting, and a structurally informed ensemble transform Kalman filter (SETKF) to embed local statistical structure into the assimilation step, assuming partial structural information such as the transformation between physical and conserved variables. Our method operates in a highly constrained environment, using only a single noisy trajectory for both training and assimilation. Numerical experiments where the unknown dynamics correspond respectively to the shallow water and Euler equations demonstrate significantly improved predictive accuracy.
The experiences of care-partners of people with epilepsy are complex and dynamic throughout the lifespan, yet their wellbeing remains insufficiently recognized within care systems. There is a public health need to examine and address the wellbeing of epilepsy care-partners. Utilizing the National Institute of Health (NIH) Whole Person Health framework, we present a narrative synthesis of often-overlooked challenges faced by epilepsy care-partners, along with potential strategies for improving their identification, measurement, support and advocacy. These considerations are essential to advancing a whole-person health approach that addresses the needs of both individuals with epilepsy and their care-partners as an interconnected dyad. We present a call to action for clinicians and systems of epilepsy care to invest in research and resources to improve health outcomes and wellbeing for care-partners of people with epilepsy; and to innovate epilepsy center care pathways that are inclusive of whole person health values and logic for epilepsy dyads.
Osteoradionecrosis (ORN) is a debilitating sequela of radiotherapy (RT) for head and neck cancer (HNC), characterized by persistent bone exposure or radiographic non-viability and profound tissue repair failure. Despite advances in radiation delivery and clinical management, ORN remains challenging to treat, with therapeutic strategies largely focused on symptom control, infection mitigation, or surgical reconstruction rather than restoration of regenerative capacity. Emerging evidence suggests that radiation-induced cellular senescence represents a central and underrecognized driver of ORN pathophysiology. Ionizing radiation disrupts the bone microenvironment through sustained vascular injury, impaired osteogenic-angiogenic coupling, chronic inflammation, and the accumulation of senescent bone-resident and stromal cells that secrete a pro-inflammatory senescence-associated secretory phenotype (SASP). Together, these changes create a hostile niche that suppresses bone remodeling and wound healing long after radiation exposure. This review synthesizes current understanding of radiation-induced injury to the bone microenvironment, critically evaluates existing classification systems and treatment strategies for ORN, and examines emerging regenerative approaches. We further propose senotherapeutic strategies targeting radiation-induced cellular senescence as a complementary and potentially transformative avenue to restore tissue repair capacity. By integrating regenerative biomaterials with targeted modulation of senescence, future therapies may move beyond palliation toward true biological resolution of ORN. Biomaterial delivery systems (e.g., injectable hydrogels, mineral composite scaffolds, and cryogel platforms) are emerging as promising vehicles for spatiotemporally controlled senotherapeutic delivery within the irradiated bone niche.
AI agents may soon become capable of autonomously completing valuable, long-horizon tasks in diverse domains. Current benchmarks either do not measure real-world tasks, or are not sufficiently difficult to meaningfully measure frontier models. To this end, we present Terminal-Bench 2.0: a carefully curated hard benchmark composed of 89 tasks in computer terminal environments inspired by problems from real workflows. Each task features a unique environment, human-written solution, and comprehensive tests for verification. We show that frontier models and agents score less than 65% on the benchmark and conduct an error analysis to identify areas for model and agent improvement. We publish the dataset and evaluation harness to assist developers and researchers in future work at tbench.ai.
Agent Skills are structured packages of procedural knowledge that augment large language model (LLM) agents at inference time. Despite rapid adoption, there is no standard way to measure whether they actually help. We present SkillsBench, a benchmark whose current inventory contains 87 tasks across 8 domains paired with curated Skills and deterministic verifiers. Our latest aggregate evaluation runs the 87-task benchmark under matched no-Skills and curated-Skills conditions for 18 model-harness configurations. Curated Skills raise the average pass rate from 33.9
