{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,18]],"date-time":"2026-03-18T02:40:20Z","timestamp":1773801620584,"version":"3.50.1"},"reference-count":0,"publisher":"Association for the Advancement of Artificial Intelligence (AAAI)","issue":"11","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["AAAI"],"abstract":"<jats:p>Retrieval Augmented Generation enhances the response accuracy of Large Language Models (LLMs) by integrating retrieval and generation modules with external knowledge, demonstrating particular strength in real-time queries and Visual Question Answering tasks. \nHowever, the effectiveness of RAG is frequently hindered by the precision of the retriever: many retrieved samples fed into the generation phase are irrelevant or misleading, posing a critical bottleneck to LLMs\u2019 performance.\nTo address this challenge, we introduce \\textbf{VaccineRAG}, a novel Chain-of-Thought-based retrieval-augmented generation dataset. \nOn one hand, VaccineRAG employs a benchmark to evaluate models using data with varying positive\/negative sample ratios, systematically exposing inherent weaknesses in current LLMs. \nOn the other hand, it enhances models\u2019 sample-discrimination capabilities by prompting LLMs to generate explicit Chain-of-Thought (CoT) analysis for each sample before producing final answers.\nFurthermore, to enhance the model\u2019s ability to learn long-sequence complex CoT content, we propose \\textbf{Partial-GRPO}. \nBy modeling the outputs of LLMs as multiple components rather than a single whole, our model can make more informed preference selections for complex sequences, thereby enhancing its capacity to learn complex CoT.\nComprehensive evaluations and ablation studies on VaccineRAG validate the effectiveness of the proposed scheme.<\/jats:p>","DOI":"10.1609\/aaai.v40i11.37876","type":"journal-article","created":{"date-parts":[[2026,3,17]],"date-time":"2026-03-17T23:47:35Z","timestamp":1773791255000},"page":"9189-9197","source":"Crossref","is-referenced-by-count":0,"title":["VaccineRAG: Boosting Multimodal Large Language Models\u2019 Immunity to Harmful RAG Samples"],"prefix":"10.1609","volume":"40","author":[{"given":"Qixin","family":"Sun","sequence":"first","affiliation":[]},{"given":"Ziqin","family":"Wang","sequence":"additional","affiliation":[]},{"given":"Hengyuan","family":"Zhao","sequence":"additional","affiliation":[]},{"given":"Yilin","family":"Li","sequence":"additional","affiliation":[]},{"given":"Kaiyou","family":"Song","sequence":"additional","affiliation":[]},{"given":"Si","family":"Liu","sequence":"additional","affiliation":[]},{"given":"Xiaolin","family":"Hu","sequence":"additional","affiliation":[]},{"given":"Qingpei","family":"Guo","sequence":"additional","affiliation":[]},{"given":"Linjiang","family":"Huang","sequence":"additional","affiliation":[]}],"member":"9382","published-online":{"date-parts":[[2026,3,14]]},"container-title":["Proceedings of the AAAI Conference on Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/ojs.aaai.org\/index.php\/AAAI\/article\/download\/37876\/41838","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/ojs.aaai.org\/index.php\/AAAI\/article\/download\/37876\/41838","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,3,17]],"date-time":"2026-03-17T23:47:36Z","timestamp":1773791256000},"score":1,"resource":{"primary":{"URL":"https:\/\/ojs.aaai.org\/index.php\/AAAI\/article\/view\/37876"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,3,14]]},"references-count":0,"journal-issue":{"issue":"11","published-online":{"date-parts":[[2026,3,17]]}},"URL":"https:\/\/doi.org\/10.1609\/aaai.v40i11.37876","relation":{},"ISSN":["2374-3468","2159-5399"],"issn-type":[{"value":"2374-3468","type":"electronic"},{"value":"2159-5399","type":"print"}],"subject":[],"published":{"date-parts":[[2026,3,14]]}}}