Improving Interactive In-Context Learning from Natural Language Feedback

Interesting paper, it reminds me how we human improving our math problem solving capability. But those efficient guys use few problems(and reasoning trajectories) to build their solving skills, how to build this effective human training process into LLM's?