Error Feedback
Error feedback, encompassing the detection, analysis, and correction of errors in various systems, is a crucial area of research aiming to improve the accuracy and reliability of models and algorithms. Current research focuses on personalized error handling, particularly in human-computer interaction and robotics, as well as developing robust methods for handling errors in data, model training, and inference, often employing techniques like contrastive learning and error majorants. These advancements have significant implications for improving the performance and trustworthiness of AI systems across diverse applications, from autonomous driving to medical diagnosis and human-robot collaboration.
Papers
Improving the realism of robotic surgery simulation through injection of learning-based estimated errors
Juan Antonio Barragan, Hisashi Ishida, Adnan Munawar, Peter Kazanzides
Advancing Tool-Augmented Large Language Models: Integrating Insights from Errors in Inference Trees
Sijia Chen, Yibo Wang, Yi-Feng Wu, Qing-Guo Chen, Zhao Xu, Weihua Luo, Kaifu Zhang, Lijun Zhang
Trial and Error: Exploration-Based Trajectory Optimization for LLM Agents
Yifan Song, Da Yin, Xiang Yue, Jie Huang, Sujian Li, Bill Yuchen Lin
Error bounds for particle gradient descent, and extensions of the log-Sobolev and Talagrand inequalities
Rocco Caprio, Juan Kuntz, Samuel Power, Adam M. Johansen