[
    {
        "id": "osp-22417",
        "type": "article-journal",
        "title": "CLIMB: Confidence-Guided Complementary Evidence for Multimodal Retrieval-Augmented Generation",
        "author": [
            {
                "family": "Gao",
                "given": "Hang"
            },
            {
                "family": "Xu",
                "given": "Wujiang"
            },
            {
                "family": "Zhang",
                "given": "Zhixing"
            },
            {
                "family": "Mei",
                "given": "Kai"
            },
            {
                "family": "Yang",
                "given": "Jingyi"
            },
            {
                "family": "Metaxas",
                "given": "Dimitris N."
            }
        ],
        "URL": "https://omanscience.com/en/articles/climb-confidence-guided-complementary-evidence-for-multimodal-retrieval-augmented-generation",
        "language": "en",
        "issued": {
            "date-parts": [
                [
                    2026
                ]
            ]
        },
        "abstract": "Multimodal large language models (MLLMs) have shown strong visual reasoning abilities, but knowledge-intensive visual question answering often requires external textual evidence beyond the image and the model's parametric knowledge. Existing multimodal RAG systems commonly rely on Top-$K$ retrieval or reranking, which may return redundant passages and provide limited control over whether an answer update is sufficiently supported by the retrieved evidence. We propose \\textit{CLIMB}, a training-free inference-time framework for multimodal RAG. CLIMB first constructs a compact complementary evidence pool using an MMR-style objective that balances query relevance and passage-level redundancy. It then performs confidence-controlled refinement within this fixed pool: an R/E/C critic scores passages by relevance, evidence specificity, and cross-modal alignment, while an evidence-grounded confidence estimator accepts an updated answer only when the estimated confidence increases. This design provides a simple stopping criterion and reduces unnecessary refinement without modifying the underlying retriever or MLLM. Experiments on Encyclopedic-VQA and InfoSeek show that CLIMB consistently improves over retrieval-augmented multimodal baselines. Ablations further indicate that complementary pooling, critic-based scoring, and iterative confidence-controlled refinement each contribute to the final performance."
    }
]