Rada Mihalcea

This project analyzes how two actions happen together in videos.

This study assesses the performance of language models on modified math problems.

A new benchmark, EmoBench, tests emotional intelligence in language models.

Strategies to enhance AI model performance with diverse and cost-effective data.

Combining causal relationships with prediction improves sentiment analysis accuracy.

This study examines the differences between real and AI-generated hotel reviews.

A study comparing inspiring social media posts from India and the UK.

A study on improving audio outputs from text prompts using preference optimization.

Examining fairness and ethics of how language models respond to users.

New method addresses regional differences in gender bias evaluation.

Research shows how prompt adjustments can enhance AI responses to diverse cultures.

Analyzing how LLMs make moral choices across languages and cultures.

Robots collaborate to discuss and share cultural insights from around the world.

Exploring a dataset focused on humor comprehension in Chinese culture.