Jisun An

We quantify the evolution of gender stereotypes in Japan from 1900 to 1998, covering the prewar–postwar transition, using a series of yearly word embeddings trained on historical text corpora. We define the gender stereotype value to measure the strength of a word’s gender association by computing the difference in cosine similarity of the word to female- versus male-related attribute words. We examine trajectories of gender stereotype across three traditionally gendered domains: Home, Work, and Politics. To provide a more granular analysis and strengthen the robustness of our findings in the Work domain, we also examine changes in gender stereotypes across 18 occupations and calculate their correlations with gender participation statistics. Our results reveal domain-specific patterns. In the Home domain, female stereotype values remain stable over time, showing no statistically significant changes. In contrast, the Work and Politics domains exhibit significant trend reversals in female stereotype values around 1945, shifting from negative prewar trends to positive postwar trends, indicating increasing female associations in public domains over time. These findings show that trends in gendered cultural discourse shifted unevenly around 1945: the observed changes in the Work and Politics domains are temporally aligned with postwar institutional transformations, including reforms under the U.S.-led Allied Occupation, while the Home domain shows more limited change. Furthermore, female stereotype values for occupations show a moderately positive correlation (r=0.364) with the proportion of women in each occupation, indicating that word-embedding-based measures of gender stereotype mirrored demographic shifts to a meaningful extent.

As AI systems increasingly engage in creative collaboration, they are often designed with the assumption that appropriate behavior will foster perceptions of equal partnership. However, it remains unclear whether partner identity continues to shape perceptions in open-ended creative tasks when behavior is held constant. To examine how perceived partner identity (AI vs. human) and partner behavior (cooperative vs. aggressive) shape perceptions and behaviors, we used collaborative drawing, a turn-based non-verbal method from art therapy. In a web-based deception study (N = 30), participants collaborated with a human confederate while being told their partner was AI or human. While cooperative behavior improved overall perceptions regardless of identity, creativity ratings consistently favored human partners despite identical behavior. Participants also adopted different collaboration strategies depending on perceived identity. These findings indicate that cooperative behavior alone is insufficient to overcome identity-based biases in creative collaboration, highlighting the structural limit of behavior-centered AI design in creative collaboration.

We investigate the performance of Large Language Model (LLM)-based zero-shot stance detection on tweets. Using FlanT5-XXL, an instruction-tuned open-source LLM, with the SemEval 2016 Tasks 6A, 6B, and P-Stance datasets, we analyze how its performance varies under different prompts and decoding strategies, as well as potential model biases. We show that the zero-shot approach can match or outperform state-of-the-art methods, including fine-tuned models. Additionally, we provide practical insights into its performance, including sensitivity to instructions and prompts, decoding strategies, prompt perplexity, and the role of negations and oppositions. We ensure that the LLM has not been trained on test datasets and identify a positivity bias that may partially explain performance differences across decoding strategies. Finally, we conduct a qualitative analysis of cases where the LLM consistently fails, uncovering questionable ground truth labels and an overconfidence in assigning a stance when none exists. In sum, we provide an in-depth case study of using an LLM for a stance detection task, which can serve as a guide for practitioners seeking to leverage LLMs for similar tasks and use cases.

Personalized AI agents rely on access to a user's digital footprint, which often includes sensitive data from private emails, chats and purchase histories. Yet this access creates a fundamental societal and privacy risk: systems lacking social-context awareness can unintentionally expose user secrets, threatening digital well-being. We introduce PrivacyBench, a benchmark with socially grounded datasets containing embedded secrets and a multi-turn conversational evaluation to measure secret preservation. Testing Retrieval-Augmented Generation (RAG) assistants reveals that they leak secrets in up to 26.56% of interactions. A privacy-aware prompt lowers leakage to 5.12%, yet this measure offers only partial mitigation. The retrieval mechanism continues to access sensitive data indiscriminately, which shifts the entire burden of privacy preservation onto the generator. This creates a single point of failure, rendering current architectures unsafe for wide-scale deployment. Our findings underscore the urgent need for structural, privacy-by-design safeguards to ensure an ethical and inclusive web for everyone.

The expansion of short-term rental platforms, such as Airbnb, has significantly disrupted local housing markets, often leading to increased rental prices and housing affordability issues. Accurately forecasting regional Airbnb market trends can thus offer critical insights for policymakers and urban planners aiming to mitigate these impacts. This study proposes a novel time-series forecasting framework to predict three key Airbnb indicators -- Revenue, Reservation Days, and Number of Reservations -- at the regional level. Using a sliding-window approach, the model forecasts trends 1 to 3 months ahead. Unlike prior studies that focus on individual listings at fixed time points, our approach constructs regional representations by integrating listing features with external contextual factors such as urban accessibility and human mobility. We convert structured tabular data into prompt-based inputs for a Large Language Model (LLM), producing comprehensive regional embeddings. These embeddings are then fed into advanced time-series models (RNN, LSTM, Transformer) to better capture complex spatio-temporal dynamics. Experiments on Seoul's Airbnb dataset show that our method reduces both average RMSE and MAE by approximately 48% compared to conventional baselines, including traditional statistical and machine learning models. Our framework not only improves forecasting accuracy but also offers practical insights for detecting oversupplied regions and supporting data-driven urban policy decisions.

We quantify the evolution of gender stereotypes in Japan from 1900 to 1999 using a series of 100 word embeddings, each trained on a corpus from a specific year. We define the gender stereotype value to measure the strength of a word's gender association by computing the difference in cosine similarity of the word to female- versus male-related attribute words. We examine trajectories of gender stereotype across three traditionally gendered domains: Home, Work, and Politics, as well as occupations. The results indicate that language-based gender stereotypes partially evolved to reflect women's increasing participation in the workplace and politics: Work and Politics domains become more strongly female-stereotyped over the years. Yet, Home also became more female-stereotyped, suggesting that women were increasingly viewed as fulfilling multiple roles such as homemakers, workers, and politicians, rather than having one role replace another. Furthermore, the strength of female stereotype for occupations positively correlate with the proportion of women in each occupation, indicating that word-embedding-based measures of gender stereotype mirrored demographic shifts to a considerable extent.

Quantifying identity fusion -- the psychological merging of self with another entity or abstract target (e.g., a religious group, political party, ideology, value, brand, belief, etc.) -- is vital for understanding a wide range of group-based human behaviors. We introduce the Cognitive Linguistic Identity Fusion Score (CLIFS), a novel metric that integrates cognitive linguistics with large language models (LLMs), which builds on implicit metaphor detection. Unlike traditional pictorial and verbal scales, which require controlled surveys or direct field contact, CLIFS delivers fully automated, scalable assessments while maintaining strong alignment with the established verbal measure. In benchmarks, CLIFS outperforms both existing automated approaches and human annotation. As a proof of concept, we apply CLIFS to violence risk assessment to demonstrate that it can improve violence risk assessment by more than 240%. Building on our identification of a new NLP task and early success, we underscore the need to develop larger, more diverse datasets that encompass additional fusion-target domains and cultural backgrounds to enhance generalizability and further advance this emerging area. CLIFS models and code are public at https://github.com/DevinW-sudo/CLIFS.