The formation of the first stars in the universe could be significantly impacted by the effects of Dark Matter (DM). Namely, if DM is in the form of Weakly Interacting Massive Particles (WIMPs), it could lead to the formation (at z similar to 25-10) of stars that are powered by DM annihilations alone, i.e. Dark Stars (DSs). Those objects can grow to become supermassive (M similar to 106M ae) and shine as bright as a galaxy (L similar to 108M ae). Using a simple 22 minimization, the first three DSs photometric candidates (i.e. JADES-GS-z11-0 , JADES-GS-z12-0 , and JADES-GS-z13-0 ) were identified by Ilie et al., (2023). Our goal is to develop tools to streamline the identification of such candidates within the rather large publicly available high redshift JWST data sets. We present here the key first step in achieving this goal: the development and implementation of a feed-forward neural network (FFNN) search for Dark Star candidates, using data from the JWST Advanced Deep Extragalactic Survey (JADES) photometric catalog. Our method reconfirms JADES-GS-z13 and JADES-GS-z11 as dark star candidates, based on the chi-squared goodness of fit test, yet it is similar to 104 times faster than the Neadler-Mead 22 minimization method used in Ilie et al., (2023). We further identify six new photometric Dark Star candidates across redshifts z similar to 9 to z similar to 14. These findings underscore the power of neural networks in modeling non-linear relationships and efficiently analyzing large-scale photometric surveys, advancing the search for Dark Stars.
Some of the first stars in the Universe might be powered by Dark Matter (DM) annihilations, rather than nuclear fusion. Those objects, i.e. Dark stars (DS), offer a unique window into understanding DM via the observational study of the first stars and their Black Hole (BH) remnants. In Mahmud et al., (2025b) (Paper I) we introduced a feedforward neural network (FFNN) trained on synthetic DS photometry in order to detect and characterize dark star photometric candidates based on data taken with the NIRCam instrument onboard the James Webb Space Telescope (JWST). In this work we develop a FFNN trained on synthetic DS spectra in order to identify spectroscopic dark star candidates in JWST's NIRSpec data. In order to validate our FFNN model we apply it to real data for the four spectroscopic Supermassive Dark Star (SMDS) candidates recently identified in Ilie et al., (2025) and reconfirm that indeed JADES-GS-z11-0, JADES-GS-z13-0, JADES-GS-z14-0, and JADES-GS-z14-1 have spectra that are consistent with those of Supermassive Dark Stars. The main advantage of our FFNN model, in comparison to the Nedler-Mead parameter estimator used in Ilie et al., (2025), is that the approach introduced here predicts parameters in milliseconds, over 10,000 times faster than the traditional method used in Ilie et al., (2025). With this in mind, the FFNN model we developed and validated in this work will be adapted for Bayesian uncertainty analyses and automatic analyses of NIRSpec publicly available data for high redshift objects. This study establishes a robust and efficient tool for probing Dark Stars and understanding their role in cosmic evolution.
Kansas City, Missouri, has seen record rent growth in recent years. This comes at a time when the city reels from a deep shortage in affordable housing, an ongoing eviction crisis, and persistent racial segregation. How is investment in new rental housing intertwined with entrenched patterns of anti-Black residential segregation and long-standing spatial inequalities? How are speculative real estate practices understood and legitimated as neutral, rational, and nonracial? This article examines these questions through ethnographic fieldwork on the everyday practices of real estate development. By examining the calculative routines and anticipatory devices deployed in the financing and designing of rental housing, the article shows how the constitution of market-rate apartment buildings into legible, predictable, and tradable financial assets is premised on reinscribing anti-Blackness onto the built environment vis-& agrave;-vis ostensibly neutral market categories and financial concepts.
Prehospital emergency medicine is a high-risk environment where time-critical decisions must be made under adverse conditions. Artificial Intelligence (AI) holds the potential to enhance patient care through data-driven decision support. This narrative review analyses current and emerging AI applications within emergency medical services and evaluates their impact on the quality dimensions defined by Donabedian (structural, process, and outcome quality). It is intended for paramedics and clinicians in the prehospital setting and aims to encourage further engagement with AI in emergency medical services to reduce cognitive load and ultimately improve patient outcomes. A review of international literature was conducted to identify AI applications in emergency medical services. The analysis was structured according to typical operational phases: emergency call interrogation, dispatch and resource allocation, operational-tactical applications, and clinical applications. Additionally, hypothetical scenarios and illustrative case studies were developed to demonstrate operational pathways and practical utility. Evaluation was guided by established healthcare quality indicators and conducted within a narrative, qualitatively integrative framework, reflecting the methodological heterogeneity of the available evidence. AI-based technologies demonstrate promising applications across all operational phases, but implementation is still mostly limited to pilot projects and local solutions. In emergency call interrogation, AI-driven speech recognition can increase the detection rate of out-of-hospital cardiac arrest by 43
Programmers are turning to AI coding assistants to answer questions about their code. Benchmarks are needed to soundly evaluate these systems and understand their performance. To enable such a study, we curate a benchmark of real-world contextualized questions derived from Github pull request comments. Out of this work, we present RubberDuckBench: a multilingual benchmark of questions about code, along with detailed rubrics for evaluating answers. We evaluate a diverse set of 20 LLMs (proprietary open-source) on answering these questions. We find that even state of the art models fail to give consistent, correct responses across the benchmark. Grok 4 (69.29
