Multimodal Scientific Claim Verification Using Large Language Models and Retrieval-Augmented Generation

This paper presents a dissertation-stage research framework for multimodal scientific claim verification using Large Language Models and Retrieval-Augmented Generation. The proposed approach extends text-centered scientific claim verification toward complete scientific PDFs by preserving layout, linking multimodal evidence units, applying hybrid retrieval, and performing staged cross-modal reasoning. The literature indicates that scientific claim verification, table reasoning, visual evidence localization, document-layout understanding, and RAG reliability are active but fragmented areas [1]-[30]. The proposed research addresses the integration gap by treating evidence provenance and uncertainty as explicit components of the verification workflow. Because this work is at Dissertation Phase-I, the manuscript deliberately reports the proposed methodology and evaluation plan rather than claiming experimental results that have not yet been obtained. Subsequent phases will implement the system, conduct controlled experiments and ablations, and report measured outcomes against the defined baselines and metrics. [11]

1. Wadden, D., Lin, S., Lo, K., Wang, L. L., van Zuylen, M., Cohan, A., and Hajishirzi, H., —Fact or Fiction: Verifying Scientific Claims,‖ Proceedings of EMNLP, pp. 7534–7550, 2020. 2. Pradeep, R., Ma, X., Nogueira, R., and Lin, J., —Scientific Claim Verification with VerT5erini,‖ Proceedings of the 12th Workshop on Health Text Mining and Information Analysis, pp. 94–103, 2021. 3. Wadden, D., Lo, K., Wang, L. L., Cohan, A., Beltagy, I., and Hajishirzi, H., —MultiVerS: Improving Scientific Claim Verification with Weak Supervision and Full-Document Context,‖ Findings of NAACL, pp. 61–76, 2022. 4. Wadden, D., Lo, K., Kuehl, B., Cohan, A., Beltagy, I., Wang, L. L., and Hajishirzi, H., —SciFact-Open: Towards Open-Domain Scientific Claim Verification,‖ Findings of EMNLP, pp. 4719–4734, 2022. 5. Lu, X., Pan, L., Liu, Q., Nakov, P., and Kan, M.-Y., —SCITAB: A Challenging Benchmark for Compositional Reasoning and Claim Verification on Scientific Tables,‖ Proceedings of EMNLP, pp. 7787–7813, 2023. 6. Akhtar, M., Subedi, N., Gupta, V., Tahmasebi, S., Cocarascu, O., and Simperl, E., —ChartCheck: Explainable Fact-Checking over Real-World Chart Images,‖ Findings of ACL, pp. 13921–13937, 2024. 7. Wang, C., Shen, Y., Kuang, Z., Cohan, A., and Zhao, Y., —SciVer: Evaluating Foundation Models for Multimodal Scientific Claim Verification,‖ Proceedings of ACL, pp. 8562–8579, 2025. 8. Lal, Y. K., Bandham, M., Hasan, M. S., Kashi, A., Koupaee, M., and Balasubramanian, N., —MuSciClaims: Multimodal Scientific Claim Verification,‖ Proceedings of IJCNLP-AACL, pp. 3285–3307, 2025. 9. Ho, X., Wu, Y.-A., Kumar, S., Xia, T. C., Boudin, F., Greiner-Petter, A., and Aizawa, A., —SciClaimEval: Cross-modal Claim Verification in Scientific Papers,‖ Proceedings of LREC, 2026. 10. Ho, X., Kumar, S., Wu, Y.-A., Boudin, F., Takasu, A., and Aizawa, A., —Table-Text Alignment: Explaining Claim Verification Against Tables in Scientific Papers,‖ Findings of EMNLP, pp. 2509–2517, 2025.