Alice Li

This study analyzes how language models recover from reasoning errors during tasks.

This system provides a scalable environment to test autonomous agents across real-world Android applications.

A study on fine-tuning computer control agents to enhance task performance.