Browse Articles

This article explored the question of when do people decide to stop gambling and further tries to extrapolate why people stop gambling at that point. Their study showed that people tend to quit gambling after 4 consecutive losses, significantly more than 1-3 consecutive losses or a win previous to quitting. They also found that participants commonly quit at a point value approximately 5 points greater than or less than their starting balance. The authors concluded that these results may be important in understanding how to cut down on excessive gambling or in creating policies that make it easier for people to disengage from gambling.

Here, the authors investigated the integration of large language models (LLMs) with drug target affinity predictors (DTAPs) to improve drug repurposing, demonstrating a significant increase in prediction accuracy, particularly with GPT-4, for psychotropic drugs and the sigma-1 receptor. This novel approach offers to potentially accelerate and reduce the cost of drug discovery by efficiently identifying new therapeutic uses for existing drugs.

Here, the authors sought to better understand how and why people experience beauty in music. They explored this fundamental aesthetic response by considering numerous emotional responses of participants to diverse musical excerpts using a 42-item Aesthetic Emotions Scale assessment. They found that the highly nuanced emotional experience of beauty in music includes positive, negative, and knowledge-related feelings.

Though current strategies in counterterrorism are somewhat effective, the Simultaneous Attack/Persistent Pursuit (SAPP) Theory may be superior alternative to current methods. The authors simulated five attack strategies (1 SAPP and 4 non-SAPP), and concluded that the SAPP model was significantly more effective in reducing the final number of terrorist attacks. This demonstrates the comparative advantage of utilizing the SAPP model, which may prove to be critical in future efforts in counterterrorism.

Palermo et al. examined the effect of group size on drinking habits of college and high school students. The authors found that both high school and college students tended to consume the most alcohol in group sizes of 4 or more, independent of how frequently they drink. They also found that the proportion of college students that drink is nearly twice the proportion of high school students that drink. This study supports previous findings that underage drinking happens in large groups and suggests that effective intervention in underage drinking would be at the group level.

This study used machine learning models to examine which factors most influenced U.S. household energy consumption in 2020 using data from 18,496 households.

Here the authors provided greater coverage of adolescent stances by investigating the political perspectives and trends of high school journalists, utilizing web scraping methods and artificial intelligence (ChatGPT-4o) to analyze over 153,000 articles. They found that high school publications exhibit lower levels of political polarization compared to mainstream media and that journalists' views, while tending to lean moderately liberal, showed no significant correlation with local voting patterns.

We conducted this research as our start-up's research that addresses the problem of biogas production in cow-dense regions like India. We hypothesized that the thermophilic temperature (45-60^oC) would increase biogas production. The production process is much faster and more abundant at temperatures around 55-60^oC.

Here the authors investigated the role of agricultural fertilizers as potential contributors to greenhouse gas emissions. In contrast to the typical investigations that consider microbiological processes, the authors considered purely chemical processes. Based on their results they found that as much as 20.41% of all CO₂ emission from land-based activities could be a result of mineral nitrogen fertilizers.

Machine translation remains a challenging area in artificial intelligence, with neural machine translation (NMT) making significant strides over the past decade but still facing hurdles, particularly in translation quality due to the reliance on expensive bilingual training data. This study explores whether large language models (LLMs), like GPT-4, can be effectively adapted for translation tasks and outperform traditional NMT systems.