Yanghua Xiao

Introducing a framework to enhance decision-making in language agents during complex tasks.

A new benchmark tests LLMs' abilities with structured data formats.

VCEval offers an automated way to assess online course effectiveness.

DetectBench evaluates LLMs on their ability to detect hidden evidence in reasoning tasks.

A novel method to fine-tune language models efficiently with fewer parameters.

A tool to identify misleading answers from large language models.

Adapting prompts to specific models improves performance in language tasks.

Research investigates how well language models understand humor in Chinese.

A new method improves meme caption generation for single and multi-image formats.

Research assesses how well LLMs generate educational questions for learning.

A novel method enhances detection and explanation of fake news.

A new framework evaluates how well language models recognize and respond to emotions.

Examining the role of emotions in enhancing language model interactions.

A new dataset and framework for generating engaging comments on Chinese videos.

This study examines how AI can help find historical analogies for current events.

BrainKing assesses language models' problem-solving skills under limited information.

Using multiple programming languages to enhance math reasoning effectively.