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

2025/08/03 by Yi Jiang, Jiang, Yi, Sendong Zhao +9
Computer Science · #Advanced Graph Neural Networks #Artificial Intelligence (cs.AI) #Computation and Language (cs.CL) #FOS: Computer and information sciences #Multimodal Machine Learning Applications #Topic Modeling

paper · pdf · doi:10.48550/arxiv.2508.01696

openalex publication_date 2025/08/03 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/31

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.

Citations

• Adaptive Contrastive Decoding in Retrieval-Augmented Generation for Handling Noisy Contexts
• Improving Retrieval Augmented Language Model with Self-Reasoning
• Synergistic Multi-Agent Framework with Trajectory Learning for Knowledge-Intensive Tasks
• InstructRAG: Instructing Retrieval-Augmented Generation via Self-Synthesized Rationales
• SuRe: Summarizing Retrievals using Answer Candidates for Open-domain QA of LLMs
• RAFT: Adapting Language Model to Domain Specific RAG
• BlendFilter: Advancing Retrieval-Augmented Large Language Models via Query Generation Blending and Knowledge Filtering
• Blinded by Generated Contexts: How Language Models Merge Generated and Retrieved Contexts When Knowledge Conflicts?
• Retrieval-Augmented Generation for Large Language Models: A Survey
• Chain-of-Note: Enhancing Robustness in Retrieval-Augmented Language Models
• CareerX: A Retrieval-Augmented Generation Framework for Personalized AI-Driven Career Guidance
• Efficient Memory Management for Large Language Model Serving with PagedAttention
• Direct Preference Optimization: Your Language Model is Secretly a Reward Model
• Trusting Your Evidence: Hallucinate Less with Context-aware Decoding
• LLaMA: Open and Efficient Foundation Language Models
• Rethinking with Retrieval: Faithful Large Language Model Inference
• Interleaving Retrieval with Chain-of-Thought Reasoning for Knowledge-Intensive Multi-Step Questions
• Generate rather than Retrieve: Large Language Models are Strong Context Generators
• Unsupervised Dense Information Retrieval with Contrastive Learning
• Constructing A Multi-hop QA Dataset for Comprehensive Evaluation of Reasoning Steps
• Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks
• Dense Passage Retrieval for Open-Domain Question Answering
• TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for\n Reading Comprehension
• When Not to Trust Language Models: Investigating Effectiveness of Parametric and Non-Parametric Memories

Related

• Synergizing RAG and Reasoning: A Systematic Review
• SIRAG: Towards Stable and Interpretable RAG with A Process-Supervised Multi-Agent Framework
• Chain of Agents: Large Language Models Collaborating on Long-Context Tasks
• MA-RAG: Multi-Agent Retrieval-Augmented Generation via Collaborative Chain-of-Thought Reasoning
• Retrieval-Generation Synergy Augmented Large Language Models