Paiheng Xu

Paiheng Xu, Louiqa Raschid, Vanessa Frias-Martinez

Social media platforms like Twitter (now X) have been pivotal in information dissemination and public engagement. The objective of our research is to analyze the effect of localized engagement on social media conversations. This study examines the impact of geographic co-location, as a proxy for localized engagement. Our research is grounded in a COVID-19 dataset. A key goal during the pandemic for public health experts was to encourage prosocial behavior that could impact local outcomes such as masking and social distancing. Given the importance of local news and guidance during COVID-19, we analyze the effect of localized engagement, between public health experts (PHEs) and the public, on social media. We analyze a Twitter Conversation dataset from January 2020 to November 2021, comprising over 19 K tweets from nearly five hundred PHEs, and 800 K replies from 350 K participants. We use a Poisson regression model to show that geo-co-location is indeed associated with higher engagement. Lexical features associated with emotion and personal experiences were more common in geo-co-located conversations. To complement our statistical analysis, we also applied a large language model (LLM)-based method to automatically generate and evaluate hypotheses; the LLM results confirm the results using lexical features. This research provides insights into how geographic co-location influences social media engagement and can inform strategies to improve public health messaging.

Yuhang Zhou, Jing Zhu, Shengyi Qian et al.

Large Language Models (LLMs) are increasingly aligned with human preferences through Reinforcement Learning from Human Feedback (RLHF). Among RLHF methods, Group Relative Policy Optimization (GRPO) has gained attention for its simplicity and strong performance, notably eliminating the need for a learned value function. However, GRPO implicitly assumes a balanced domain distribution and uniform semantic alignment across groups, assumptions that rarely hold in real-world datasets. When applied to multi-domain, imbalanced data, GRPO disproportionately optimizes for dominant domains, neglecting underrepresented ones and resulting in poor generalization and fairness. We propose Domain-Informed Self-Consistency Policy Optimization (DISCO), a principled extension to GRPO that addresses inter-group imbalance with two key innovations. Domain-aware reward scaling counteracts frequency bias by reweighting optimization based on domain prevalence. Difficulty-aware reward scaling leverages prompt-level self-consistency to identify and prioritize uncertain prompts that offer greater learning value. Together, these strategies promote more equitable and effective policy learning across domains. Extensive experiments across multiple LLMs and skewed training distributions show that DISCO improves generalization, outperforms existing GRPO variants by 5% on Qwen3 models, and sets new state-of-the-art results on multi-domain alignment benchmarks. Our code and data are available at https://github.com/Tonyzhou98/disco_grpo.