Agent-Pro: Learning to Evolve via Policy-Level Reflection and Optimization

arXiv:2402.17574v1 Announce Type: cross
Abstract: Large Language Models exhibit robust problem-solving capabilities for diverse tasks. However, most LLM-based agents are designed as specific task solvers with sophisticated prompt engineering, rather than agents capable of learning and evolving through interactions. These task solvers necessitate manually crafted prompts to inform task rules and regulate LLM behaviors, inherently incapacitating to address complex dynamic scenarios e.g., large interactive games. In light of this, we propose Agent-Pro: an LLM-based Agent with Policy-level Reflection and Optimization …

abstract agent agents arxiv capabilities cs.ai cs.cl diverse engineering interactions language language models large language large language models llm optimization policy problem-solving prompt prompts robust rules tasks through type via