The Graduate Center, CUNY,City University of New York

Delay lines are essential components in photonic systems, facilitating data processing, storage, interconnects, and generally enhancing light-matter interactions. They rely on slow light phenomena, which can be realized in integrated photonic platforms such as coupled-resonator optical waveguides, photonic crystals and optical waveguides. In practice, delay lines require robustness, broadband operation and small footprints. Yet, typically, these features need to be traded against each other at the design stage due to constraints stemming from passivity, linearity and time-invariance. Here, we demonstrate that compactness, robustness and bandwidth can be simultaneously achieved by routing light in the synthetic frequency dimension along topological edge states, achieving delay-bandwidth products far beyond conventional limits combined with inherent robustness.

Active learning through interactive exploration significantly enhances student engagement and understanding of chemistry. This educational activity demonstrates Principal Component Analysis (PCA) and Partial-Least-Square-Discriminant Analysis (PLS-DA), two foundational machine learning techniques widely applied in contemporary research. Interactive Python-based programs offer accessible educational platforms for students exploring chemical data, requiring no prior programming experience. This application allows learners to actively engage in feature exploration and dimensionality reduction processes, applied to clustering and classifying binary AB equiatomic solid state compounds. Students can actively select and modify chemical and physical features, observing in real time how these choices impact the effectiveness of the PCA and PLS-DA clustering models. Initially, PCA enables unsupervised visualization of natural clustering and correlations among compounds without prior labeling. Subsequently, by employing PLS-DA, students develop supervised models capable of predicting crystal structures, explicitly illustrating supervised versus unsupervised learning paradigms. The activity highlights the importance of explainability in machine learning models rather than operating the models as a '' black box ''. Beyond learning fundamental concepts, the activity encourages students to participate in genuine exploratory processes, mirroring the investigative approaches historically utilized by researchers and practiced today. By experimenting freely with data sets and computational methods, students experience firsthand the iterative nature of scientific discovery, fostering deeper insight into both chemical informatics and the broader research methodology.

Spatial transcriptomics (ST) unlocks potential for studying gene functions in processes that depend on orchestration of transcription across space. However, analysis tools for ST remain aimed at data exploration, with few resources for hypothesis testing. What's missing is a way to test whether a factor of interest affects functionally relevant parameters of a gene's spatial distribution. We present a tool to fill this gap, which we call a warped sigmoidal Poisson-process mixed-effects (WSP, pronounced "wisp") model. WSP models are the first ST tool allowing researchers to test critical questions without bespoke preprocessing pipelines for identifying key spatial parameters. By aligning coordinates to an axis of interest and letting a likelihood-based regression find between-group effects on expression rates and boundaries, WSP models replace error-prone manual preprocessing with minimally biased hypothesis testing. After introducing WSP models, we demonstrate their statistical validity using semi-synthetic simulated data and their ability to test for effects by applying them to MERFISH data from mouse somatosensory cortex and bulk sequencing data from mouse liver lobules with extrapolated spatial coordinates. Together, these validations and applications demonstrate that WSP models offer a practical and statistically rigorous approach to quantifying and testing for effects on spatial variation in transcriptomic data.

A persistent puzzle throughout the cognitive and social sciences is how people manage to learn social conventions from the sparse and noisy behavioral data of diverse actors, without explicit instruction. Here, we show that the dominant theories of social learning perform poorly at capturing how individuals learn conventions in coordination experiments that task them with matching their behaviors while interacting in social networks. Across experiments, participants' choices systematically deviate from both imitation and optimization. Instead, they follow a categorical, two-stage learning process: they behave probabilistically until they acquire enough information about each other to trigger a mental threshold and then their choices stabilize. We effectively estimate this threshold using the Tolerance Principle (TP), a parameter-free equation first developed to model how children learn rules in language. We show that threshold-based agents produce social learning that is more accurate than imitating and optimizing agents, while also providing a better model of how a critical mass of dissenters can overturn conventions. The superior performance of our model holds when comparing against a variety of optimization approaches, including Bayesian inference. Furthermore, in a preregistered dyadic experiment requiring people to infer non-linguistic behavioral patterns amid controlled levels of noise in observed signals, TP outperforms all other models at reproducing learning rates among human participants. These findings offer compelling evidence that a simple, mathematical threshold underlies individual and social learning, from grammatical rules to behavioral conventions.

We study how news shocks to TFP, foreign interest rates (R*), and commodity terms of trade (CToT) affect output variance across countries. The difference in variance share of output explained by news shocks between developing and developed countries is: I) Negligible for news shocks to TFP. II) Positive for news shocks to R* (7 pp) and CToT (9.6 pp). Less financially developed countries are more influenced by news shocks to R*, and countries with less developed derivatives markets are more impacted by news shocks to CToT. A two-sector small open economy model with non-financially included households and hedging in commodity prices explains our findings.