CoCoA: Collaborative Chain-of-Agents for Parametric-Retrieved Knowledge Synergy

TL;DR

This paper introduces CoCoA, a multi-agent framework that improves the synergy between an LLM's internal knowledge and external retrievals, leading to better performance in knowledge-intensive tasks.

Contribution

It proposes a novel multi-agent RAG framework and a long-chain training strategy to explicitly enhance knowledge synergy in LLMs.

Findings

Outperforms existing methods in open-domain QA

Enhances multi-hop reasoning capabilities

Improves knowledge integration effectiveness

Abstract

Retrieval-Augmented Generation (RAG) enhances Large Language Models (LLMs), especially for knowledge-intensive tasks. Despite its advantages, current RAG methods often struggle to fully exploit knowledge during generation. In particular, the synergy between the model's internal parametric knowledge and external retrieved knowledge remains limited. Retrieved contents may sometimes mislead generation, while certain generated content can guide the model toward more accurate outputs. In this work, we propose Collaborative Chain-of-Agents, a framework designed to enhance explicitly synergy over both parametric and retrieved knowledge. Specifically, we first introduce CoCoA-zero, a multi-agent RAG framework that first performs conditional knowledge induction and then reasons answers. Building on this, we develop CoCoA, a long-chain training strategy that synthesizes extended multi-agent reasoning trajectories from CoCoA-zero to fine-tune the LLM. This strategy enhances the model's capability to explicitly integrate and jointly leverage parametric and retrieved knowledge. Experimental results demonstrate the superiority of CoCoA in open-domain QA and multi-hop QA.