Optimal sexual passion outcomes were found at both higher and lower levels of religiosity, whereas mid-level religiosity was associated with the less beneficial sexual passion outcomes

Religious Piety and Sexual Passion: What Is the Connection? Rebecca W. Clarke, Chelom E. Leavitt, Dean M. Busby. Journal of Sex & Marital Therapy, Sep 19 2021. https://doi.org/10.1080/0092623X.2021.1979702

Abstract: The association between religiosity and the construct of sexual passion was examined to see whether religiosity is associated with harmonious, inhibited, and obsessive sexual passion styles. Using multiple regression and checking for interactions between religiosity, gender, broad categories of religion (Catholic, other Christian, other religious, nonreligious), and the three sexual passion styles, the following associations were found: Religiosity had a significant curvilinear relationship to all three sexual passion styles, suggesting an overlap in levels of religiosity and the construct of sexual passion. Optimal sexual passion outcomes were found at both higher and lower levels of religiosity, whereas mid-level religiosity was associated with the less beneficial sexual passion outcomes. Religious men were significantly more obsessively passionate than religious women, and religious men and women were similarly high on levels of harmonious and low on inhibited sexual passion. There were no significant interactions between religiosity, broad categories of religion, and sexual passion styles. Understanding how religion and sexual passion are associated could be useful for applied researchers as well as those who work with religious individuals who want to help these individuals develop beneficial patterns of sexual passion.

How truthful is GPT-3? A benchmark for language models

TruthfulQA: Measuring How Models Mimic Human Falsehoods. Stephanie Lin, Jacob Hilton, Owain Evans. arXiv, Sep 8 2021. TruthfulQA: Measuring How Models Mimic Human Falsehoods

Abstract: We propose a benchmark to measure whether a language model is truthful in generating answers to questions. The benchmark comprises 817 questions that span 38 categories, including health, law, finance and politics. We crafted questions that some humans would answer falsely due to a false belief or misconception. To perform well, models must avoid generating false answers learned from imitating human texts. We tested GPT-3, GPT-Neo/J, GPT-2 and a T5-based model. The best model was truthful on 58% of questions, while human performance was 94%. Models generated many false answers that mimic popular misconceptions and have the potential to deceive humans. The largest models were generally the least truthful. For example, the 6B-parameter GPT-J model was 17% less truthful than its 125M-parameter counterpart. This contrasts with other NLP tasks, where performance improves with model size. However, this result is expected if false answers are learned from the training distribution. We suggest that scaling up models alone is less promising for improving truthfulness than fine-tuning using training objectives other than imitation of text from the web.

Summarized: How truthful is GPT-3? A benchmark for language models. Owain Evans. Sep 16 2021. https://www.alignmentforum.org/posts/PF58wEdztZFX2dSue/how-truthful-is-gpt-3-a-benchmark-for-language-models

3. Larger models are less truthful. 
Across different model families, the largest models were generally less truthful (Figure 2). This “inverse scaling” trend contrasts with most tasks in NLP, where performance improves with model size. For example, the 6B-parameter GPT-J model was 17% less truthful than its 125M-parameter counterpart. One explanation of this result is that larger models produce more imitative falsehoods because they are better at learning the training distribution. Another explanation is that our questions adversarially exploit weaknesses in larger models not arising from imitation of the training distribution. We ran experiments aimed to tease apart these explanations.

An attractiveness judgment for one face part can be highly predictive of the attractiveness of the whole face or the other parts

Predicting attractiveness from face parts reveals multiple covarying cues. Chang Hong Liu,Andrew W. Young, Jiaxin Li, Xinran Tian, Wenfeng Chen. British Journal of Psychology, September 20 2021. https://doi.org/10.1111/bjop.12532

Abstract: In most studies of facial attractiveness perception, judgments are based on the whole face images. Here we investigated how attractiveness judgments from parts of faces compare to perceived attractiveness of the whole face, and to each other. We manipulated the extent and regions of occlusion, where either the left/right or the top/bottom half of the face was occluded. We also further segmented the face into relatively small horizontal regions involving the forehead, eyes, nose, or mouth. The results demonstrated the correlated nature of face regions, such that an attractiveness judgment for one face part can be highly predictive of the attractiveness of the whole face or the other parts. The left/right half of the face created more accurate predictions than the top/bottom half. Judgments involving a larger area of the face (i.e., left/right or top/bottom halves) produced more accurate predictions than those derived from smaller regions, such as the eyes or the mouth alone, but even the smallest and most featureless region investigated (the forehead) provided useful information. The correlated nature of the attractiveness of face parts shows that perceived attractiveness is determined by multiple covarying cues that the visual system can exploit to determine attractiveness from a single glance.

The evidence that COVID-19 infection causes or impacts Erectile Dysfunction is compelling

The Epidemic of COVID-19-Related Erectile Dysfunction: A Scoping Review and Health Care Perspective. Tung-Chin Hsieh et al. Sexual Medicine Reviews, September 20 2021. https://doi.org/10.1016/j.sxmr.2021.09.002

Abstract

Introduction: COVID-19 infection is expected to be associated with an increased likelihood of erectile dysfunction (ED). Considering the high transmissibility of COVID-19, ED may be a concerning consequence for a large segment of the population.

Aims: To (1) summarize existing published evidence for the impact of COVID-19 on the prevalence, severity, treatment, and management of ED; and (2) identify health-related trends in the emerging literature and identify gaps in the existing research literature and make recommendations for future research needs in the area.

Methods: A scoping literature search was conducted on April 27, 2021. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Extension for Scoping Reviews (PRISMA-ScR) checklist was followed. The literature search was performed in PubMed using the terms: COVID-19, erectile, sexual, and dysfunction. A total of 693 publications were screened for relevance. Studies were appraised for their level of evidence based on study design and the rigor of methodology.

Results: The evidence that COVID-19 infection causes or impacts ED is compelling. Four topics emerged regarding the nature of the association between COVID-19 and ED: (1) the biological impact of COVID-19 infection on ED; (2) the mental health impact of COVID-19 on ED; (3) the impact of COVID-19 on the management of ED and access to ED treatment; and (4) health disparities and the impact of COVID-19 on ED. Long-term and well-designed studies are needed to clarify the extent of the impact of COVID-19 on ED. The pandemic exposed several vulnerabilities within worldwide healthcare and social systems.

Conclusion: COVID-19 has a uniquely harmful impact on men's health and erectile function through biological, mental health, and healthcare access mechanisms. As the pandemic wanes, strategies to identify long-term effects and additional health care support may be needed to adequately mitigate the impact of COVID-19 on men's health.