Columbia College is the oldest undergraduate college of Columbia University, situated on the university's main campus in Morningside Heights in the borough of Manhattan in New York City. It was founded by the Church of England in 1754 as King's College, receiving a royal charter from King George II of Great Britain. It is the oldest institution of higher learning in the state of New York and the fifth oldest in the United States. As of 2020, Columbia is ranked as the third best college in the United States by U.S. News and World Report after only Princeton and Harvard and is among the most prestigious in the world. The college is distinctive for its comprehensive Core Curriculum and is among the most selective colleges in its admissions. The college has produced many distinguished alumni, including high-ranking politicians, renowned scholars, and business leaders.
Large language models (LLMs) have shown promise on summarization tasks, but they often produce hallucinations, which are unsupported or incorrect statements that limit their reliability in specialized healthcare applications. We introduce (), an inference-time method that leverages hallucination detectors to guide iterative summary revisions toward factual corrections. Building on this, we propose for Preference Learning (), which converts detector-guided refinement trajectories into preference pairs for model finetuning. Extensive experiments show that our methods substantially reduce hallucinations for Llama and Gemma models in summarizing real-world clinical notes from . For example, reduces 24% and reduces 48% hallucinations in Llama-3.1-8B-Instruct. Importantly, both methods preserve summary fluency, coherence, and relevance according to human expert and LLM-Jury evaluations. Together, these results demonstrate that detection-informed refinement and preference learning offer an automated solution for improving factual faithfulness in clinical summarization.
This article asks what the opportunities for and constraints on labor organizing in San Diego's logistics sector are. Using Amazon warehouse workers and XPO Logistics truck drivers as illustrative examples the paper uses the power resources approach to assess the associational, institutional, societal, structural, and disruptive power workers have at their disposal in the region's logistics sector. Drawn from a broader project on the San Diego-Tijuana region's logistics sector that uses 2 years of fieldwork, 25 semi-structured interviews, oral history accounts, quantitative data, and primary document analysis it is argued that while structural power is high due to occupying an important position along the supply chain amidst the ongoing global trade war, associational power and institutional power are low due to the precarious citizenship status of many workers in the region and the independent contractor status of truckers. However, there are some exceptions associated with lower rates of turnover due to the transnational nature of work and social reproduction as well as the accompanying social ties. Furthermore, it is argued that building societal power with immigrant rights and legal aid organizations is important to building the disruptive capacity of workers.
Canadian post-secondary institutions have seen a significant increase in international students over the past few years due to their mosaic pattern of multiculturalism. Canadian institutions attract many students from various countries, who desire to study in a diverse and inclusive academic environment and seek better quality education, greater job opportunities, and a higher standard of living. However, international students face challenges integrating into the foreign education system, which can have deleterious effects on their physical and mental health. This chapter synthesizes the learning and applications of available research towards the well-being of international students in post-secondary institutions in Canada, with a focus on student engagement. It analyzes the challenges faced by international students, explores the reasons for underutilization of support services by students, and examines how using a culturally sensitive approach in existing campus mental health programs helps improve accessibility for international students in Canadian institutions.
Effective decisions require not only choosing between alternative options but selecting evidence that is informative for those options. Humans often deviate from strategies that maximize information gain, but the mechanisms of these deviations are not understood. We tested adult participants in a task in which they chose which question to inspect to prepare for a test. Participants received two alternative questions and could identify the more informative question based on its uncertainty (Unc) and probability of appearing at test (PTest). Despite the simplicity and transparency of the task, participants systematically deviated from the normative strategy. Computational modeling showed that most participants underweighted Unc and PTest and relied more strongly on PTest vs Unc. Strikingly, a substantial minority systematically preferred theleast-informative question, suggesting that they incorrectly reasoned about how uncertainty impacted informativeness. Moreover, in the absence of external feedback, sub-optimal strategies were amplified over time. The findings highlight the importance of abstract reasoning, task understanding, and self-reinforcing learning in constraining human information demand.
