{"paper":{"title":"Retrieval-Augmented Generation for AI-Generated Content: A Survey","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"RAG integrates retrieval into AI-generated content to pull relevant data and raise accuracy plus robustness.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bin Cui, Fangcheng Fu, Hailin Zhang, Jie Jiang, Ling Yang, Penghao Zhao, Qinhan Yu, Wentao Zhang, Yunteng Geng, Zhengren Wang","submitted_at":"2024-02-29T18:59:01Z","abstract_excerpt":"Advancements in model algorithms, the growth of foundational models, and access to high-quality datasets have propelled the evolution of Artificial Intelligence Generated Content (AIGC). Despite its notable successes, AIGC still faces hurdles such as updating knowledge, handling long-tail data, mitigating data leakage, and managing high training and inference costs. Retrieval-Augmented Generation (RAG) has recently emerged as a paradigm to address such challenges. In particular, RAG introduces the information retrieval process, which enhances the generation process by retrieving relevant objec"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"RAG introduces the information retrieval process, which enhances the generation process by retrieving relevant objects from available data stores, leading to higher accuracy and better robustness.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the collected literature and proposed classification of augmentation methodologies comprehensively represent the space of RAG-AIGC integrations without significant omissions or overlaps.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"A survey classifying RAG foundations for AIGC, summarizing enhancements, cross-modal applications, benchmarks, limitations, and future directions.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"RAG integrates retrieval into AI-generated content to pull relevant data and raise accuracy plus robustness.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"33df38043a3987f6ff2bfb5bdad0a8c61c37a6110ce8d99da426a5509212ff65"},"source":{"id":"2402.19473","kind":"arxiv","version":6},"verdict":{"id":"df559452-f8ee-414b-ac30-e96b2a61ab14","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T13:27:39.327825Z","strongest_claim":"RAG introduces the information retrieval process, which enhances the generation process by retrieving relevant objects from available data stores, leading to higher accuracy and better robustness.","one_line_summary":"A survey classifying RAG foundations for AIGC, summarizing enhancements, cross-modal applications, benchmarks, limitations, and future directions.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the collected literature and proposed classification of augmentation methodologies comprehensively represent the space of RAG-AIGC integrations without significant omissions or overlaps.","pith_extraction_headline":"RAG integrates retrieval into AI-generated content to pull relevant data and raise accuracy plus robustness."},"references":{"count":298,"sample":[{"doi":"","year":2020,"title":"Language models are few-shot learners,","work_id":"0bea6c6b-769a-474f-bb43-fe4d93230958","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2021,"title":"Evaluating Large Language Models Trained on Code","work_id":"042493e9-b26f-4b4e-bbde-382072ca9b08","ref_index":2,"cited_arxiv_id":"2107.03374","is_internal_anchor":true},{"doi":"","year":2023,"title":"GPT-4 Technical Report","work_id":"b928e041-6991-4c08-8c81-0359e4097c7b","ref_index":3,"cited_arxiv_id":"2303.08774","is_internal_anchor":true},{"doi":"","year":2023,"title":"LLaMA: Open and Efficient Foundation Language Models","work_id":"c018fc23-6f3f-4035-9d02-28a2173b2b9d","ref_index":4,"cited_arxiv_id":"2302.13971","is_internal_anchor":true},{"doi":"","year":2023,"title":"Llama 2: Open Foundation and Fine-Tuned Chat Models","work_id":"68a5177f-d644-44c1-bd4f-4e5278c22f5d","ref_index":5,"cited_arxiv_id":"2307.09288","is_internal_anchor":true}],"resolved_work":298,"snapshot_sha256":"5db19b5ea030054dddd3b827cb80c2b8d079f39053cbb77629a66f59da93d4b5","internal_anchors":16},"formal_canon":{"evidence_count":2,"snapshot_sha256":"7c414884d81b2eebfbe30d23ce7c3c38ebf69e740d0eba8efb5e59f6f54488fa"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}