C-3PO: Compact Plug-and-Play Proxy Optimization to Achieve Human-like Retrieval-Augmented Generation

TL;DR

C-3PO introduces a multi-agent proxy framework that improves retrieval-augmented generation by enabling better communication and coordination between retrievers and language models without modifying their core components.

Contribution

The paper presents a novel multi-agent proxy system that enhances RAG performance through collaborative optimization and reinforcement learning-based coordination, without altering existing models.

Findings

Significantly improves RAG performance in various scenarios.

Maintains plug-and-play flexibility and generalization.

Effective multi-agent coordination via tree-structured rollout.

Abstract

Retrieval-augmented generation (RAG) systems face a fundamental challenge in aligning independently developed retrievers and large language models (LLMs). Existing approaches typically involve modifying either component or introducing simple intermediate modules, resulting in practical limitations and sub-optimal performance. Inspired by human search behavior -- typically involving a back-and-forth process of proposing search queries and reviewing documents, we propose C-3PO, a proxy-centric framework that facilitates communication between retrievers and LLMs through a lightweight multi-agent system. Our framework implements three specialized agents that collaboratively optimize the entire RAG pipeline without altering the retriever and LLMs. These agents work together to assess the need for retrieval, generate effective queries, and select information suitable for the LLMs. To enable effective multi-agent coordination, we develop a tree-structured rollout approach for reward credit assignment in reinforcement learning. Extensive experiments in both in-domain and out-of-distribution scenarios demonstrate that C-3PO significantly enhances RAG performance while maintaining plug-and-play flexibility and superior generalization capabilities.