Multi-Agent Systems (MAS) are commonly used to improve reasoning diversity and robustness by simulating interactions among agents with distinct roles. However, prior work often entangles the contribution of the multi-agent architecture with that of prompt conditioning, making the source of observed diversity gains unclear. We address this confound with a controlled study on divergent thinking tasks, using identical prompt conditioning for MAS and single agent baseline. Under these matched conditions, single agent setups consistently outperform multi-agent systems in semantic diversity. We attribute this gap to information visibility: parallel agents often converge on overlapping ideas, whereas a single agent model can condition on its own generation to avoid redundancy. We further find that a Multi-Output strategy, which prompts a single agent to produce multiple responses within a single inference pass, achieves the highest diversity without degrading logical validity. Together, these results point to a more efficient and effective way to expand diversity, with implications for the design of more efficient agentic frameworks.

Loneliness, a significant public health concern, is closely connected to both physical and mental well-being. Hence, detection and intervention for individuals experiencing loneliness are crucial. Identifying loneliness in text is straightforward when it is explicitly stated but challenging when it is implicit. Detecting implicit loneliness requires a manually annotated dataset because whereas explicit loneliness can be detected using keywords, implicit loneliness cannot be. However, there are no freely available datasets with clear annotation guidelines for implicit loneliness. In this study, we construct a freely accessible Japanese loneliness dataset with annotation guidelines grounded in the psychological definition of loneliness. This dataset covers loneliness intensity and the contributing factors of loneliness. We train two models to classify whether loneliness is expressed and the intensity of loneliness. The model classifying loneliness versus non-loneliness achieves an F1-score of 0.833, but the model for identifying the intensity of loneliness has a low F1-score of 0.400, which is likely due to label imbalance and a shortage of a certain label in the dataset. We validate performance in another domain, specifically X (formerly Twitter), and observe a decrease. In addition, we propose improvement suggestions for domain adaptation.

Measuring happiness as a determinant of well-being is increasingly recognized as crucial. While previous studies have utilized free-text descriptions to estimate happiness on a broad scale, limited research has focused on tracking individual fluctuations in happiness over time owing to the challenges associated with longitudinal data collection. This study addresses this issue by obtaining longitudinal data from two workplaces over two and six months respectively.Subsequently, the data is used to construct a happiness estimation model and assess individual happiness levels.Evaluation of the model performance using correlation coefficients shows variability in the correlation values among individuals.Notably, the model performs satisfactorily in estimating 9 of the 11 users’ happiness scores, with a correlation coefficient of 0.4 or higher. To investigate the factors affecting the model performance, we examine the relationship between the model performance and variables such as sentence length, lexical diversity, and personality traits. Correlations are observed between these features and model performance.