Tingzhang Zhao

DORA: Dynamic Optimization Prompt for Continuous Reflection of LLM-based Agent

Autonomous agents powered by large language models (LLMs) hold significant potential across various domains. The Reflection framework is designed to help agents learn from past mistakes in complex tasks. While previous research has shown that reflection can enhance performance, our investigation rev…