Evaluating Context Sensitivity and Cultural Bias in Large Language Models
High-Context vs. Low-Context Sensitivity in LLMs
Python
Research
Project Overview
Role | Researcher |
Timeline | 1 week (2025) |
Tools Used | Python, Scikit Learn, Plotly |
Link |
Background
Project observing whether large Language Models assign significantly lower surprisal to low-context speech acts compared to high-context equivalents. Additionally looks at whether multilingual models perform better at high-context tasks.