{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,7]],"date-time":"2026-05-07T11:15:17Z","timestamp":1778152517354,"version":"3.51.4"},"reference-count":42,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2026,3,23]],"date-time":"2026-03-23T00:00:00Z","timestamp":1774224000000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0"},{"start":{"date-parts":[[2026,5,7]],"date-time":"2026-05-07T00:00:00Z","timestamp":1778112000000},"content-version":"vor","delay-in-days":45,"URL":"https:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0"}],"funder":[{"name":"the Sichuan Transportation Science and Technology Program","award":["2018-ZL-02"],"award-info":[{"award-number":["2018-ZL-02"]}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Sci Rep"],"DOI":"10.1038\/s41598-026-44935-1","type":"journal-article","created":{"date-parts":[[2026,3,23]],"date-time":"2026-03-23T17:18:24Z","timestamp":1774286304000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Attention re-alignment in multimodal large language models via intermediate-layer guidance"],"prefix":"10.1038","volume":"16","author":[{"given":"Yanming","family":"Chen","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Pandong","family":"Wang","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Guofeng","family":"Qin","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Wei","family":"Wu","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Ming","family":"Chen","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yongtao","family":"Hao","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"297","published-online":{"date-parts":[[2026,3,23]]},"reference":[{"key":"44935_CR1","unstructured":"Team, K. et al. Kimi-VL technical report. arXiv preprint arXiv:2504.07491 (2025)."},{"key":"44935_CR2","unstructured":"Chen, J. et al. BLIP3-o: a family of fully open unified multimodal models-architecture, training and dataset. arXiv preprint arXiv:2505.09568 (2025)."},{"key":"44935_CR3","unstructured":"Bai, S. et al. Qwen2. 5-vl technical report. arXiv preprint arXiv:2502.13923 (2025)."},{"key":"44935_CR4","unstructured":"Wu, Z. et al. Deepseek-vl2: mixture-of-experts vision-language models for advanced multimodal understanding. arXiv preprint arXiv:2412.10302 (2024)."},{"key":"44935_CR5","unstructured":"Li, F. et al. Llava-next-interleave: tackling multi-image, video, and 3D in large multimodal models. arXiv preprint arXiv:2407.07895 (2024)."},{"key":"44935_CR6","doi-asserted-by":"crossref","unstructured":"Huang, Q. et al. Opera: Alleviating hallucination in multi-modal large language models via over-trust penalty and retrospection-allocation. In Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition 13418\u201313427 (2024).","DOI":"10.1109\/CVPR52733.2024.01274"},{"key":"44935_CR7","unstructured":"Zhang, Y.-F. et al. Debiasing multimodal large language models. arXiv preprint arXiv:2403.05262 (2024)."},{"key":"44935_CR8","unstructured":"Zhou, G. et al. Mitigating modality prior-induced hallucinations in multimodal large language models via deciphering attention causality. arXiv preprint arXiv:2410.04780 (2024)."},{"key":"44935_CR9","unstructured":"Zhang, J., Khayatkhoei, M., Chhikara, P. & Ilievski, F. Mllms know where to look: training-free perception of small visual details with multimodal llms. arXiv preprint arXiv:2502.17422 (2025)."},{"key":"44935_CR10","unstructured":"Wang, C. et al. Mllm can see? Dynamic correction decoding for hallucination mitigation. arXiv preprint arXiv:2410.11779 (2024)."},{"key":"44935_CR11","doi-asserted-by":"crossref","unstructured":"Zhang, C. et al. Self-correcting decoding with generative feedback for mitigating hallucinations in large vision-language models. arXiv preprint arXiv:2502.06130 (2025).","DOI":"10.1109\/EIECS67708.2025.11283452"},{"key":"44935_CR12","first-page":"120578","volume":"37","author":"Y Liang","year":"2024","unstructured":"Liang, Y. et al. Unleashing region understanding in intermediate layers for MLLM-based referring expression generation. Adv. Neural. Inf. Process. Syst. 37 120578\u2013120601 (2024).","journal-title":"Adv. Neural. Inf. Process. Syst."},{"key":"44935_CR13","unstructured":"Zou, X. et al. Look twice before you answer: memory-space visual retracing for hallucination mitigation in multimodal large language models. arXiv preprint arXiv:2410.03577 (2024)."},{"key":"44935_CR14","doi-asserted-by":"crossref","unstructured":"Liu, S., Zheng, K. & Chen, W. Paying more attention to image: a training-free method for alleviating hallucination in lvlms, 2024. arXiv:abs\/2407.21771 (2024).","DOI":"10.1007\/978-3-031-73010-8_8"},{"key":"44935_CR15","doi-asserted-by":"crossref","unstructured":"Fu, Y., Xie, R., Sun, X., Kang, Z. & Li, X. Mitigating hallucination in multimodal large language model via hallucination-targeted direct preference optimization. arXiv preprint arXiv:2411.10436 (2024).","DOI":"10.18653\/v1\/2025.findings-acl.850"},{"key":"44935_CR16","unstructured":"Xing, X. et al. Where do large vision-language models look at when answering questions? arXiv preprint arXiv:2503.13891 (2025)."},{"key":"44935_CR17","unstructured":"Stan, G. B.\u00a0M. et al. Lvlm-interpret: an interpretability tool for large vision-language models. arXiv preprint arXiv:2404.03118 (2024)."},{"key":"44935_CR18","unstructured":"Giulivi, L. & Boracchi, G. Explaining multi-modal large language models by analyzing their vision perception. arXiv preprint arXiv:2405.14612 (2024)."},{"key":"44935_CR19","unstructured":"Touvron, H. et al. Llama: open and efficient foundation language models. arXiv preprint arXiv:2302.13971 (2023)."},{"key":"44935_CR20","doi-asserted-by":"publisher","first-page":"nwae403","DOI":"10.1093\/nsr\/nwae403","volume":"11","author":"S Yin","year":"2024","unstructured":"Yin, S. et al. A survey on multimodal large language models. Natl. Sci. Rev. 11, nwae403 (2024).","journal-title":"Natl. Sci. Rev."},{"key":"44935_CR21","doi-asserted-by":"publisher","first-page":"23716","DOI":"10.52202\/068431-1723","volume":"35","author":"J-B Alayrac","year":"2022","unstructured":"Alayrac, J.-B. et al. Flamingo: a visual language model for few-shot learning. Adv. Neural. Inf. Process. Syst. 35, 23716\u201323736 (2022).","journal-title":"Adv. Neural. Inf. Process. Syst."},{"key":"44935_CR22","unstructured":"Li, J., Li, D., Savarese, S. & Hoi, S. Blip-2: bootstrapping language-image pre-training with frozen image encoders and large language models. In International Conference on Machine Learning 19730\u201319742 (PMLR, 2023)."},{"key":"44935_CR23","first-page":"34892","volume":"36","author":"H Liu","year":"2023","unstructured":"Liu, H., Li, C., Wu, Q. & Lee, Y. J. Visual instruction tuning. Adv. Neural. Inf. Process. Syst. 36, 34892\u201334916 (2023).","journal-title":"Adv. Neural. Inf. Process. Syst."},{"key":"44935_CR24","unstructured":"Zhu, D., Chen, J., Shen, X., Li, X. & Elhoseiny, M. Minigpt-4: enhancing vision-language understanding with advanced large language models. arXiv preprint arXiv:2304.10592 (2023)."},{"key":"44935_CR25","doi-asserted-by":"crossref","unstructured":"Sapkota, R. & Karkee, M. Object detection with multimodal large vision-language models: an in-depth review. Available at SSRN 5233953 (2025).","DOI":"10.2139\/ssrn.5233953"},{"key":"44935_CR26","doi-asserted-by":"publisher","DOI":"10.1016\/j.inffus.2025.103121","volume":"120","author":"G Khvatskii","year":"2025","unstructured":"Khvatskii, G. et al. Do multimodal large language models understand welding?. Inf. Fusion 120, 103121 (2025).","journal-title":"Inf. Fusion"},{"key":"44935_CR27","doi-asserted-by":"publisher","DOI":"10.1016\/j.patcog.2024.110898","volume":"157","author":"Y Su","year":"2025","unstructured":"Su, Y., Tan, Y., An, S., Xing, M. & Feng, Z. Semantic-driven dual consistency learning for weakly supervised video anomaly detection. Pattern Recognit. 157, 110898 (2025).","journal-title":"Pattern Recognit."},{"key":"44935_CR28","unstructured":"Arif, K. H.\u00a0I., Dip, S.\u00a0A., Hussain, K., Zhang, L. & Thomas, C. Paint: Paying attention to informed tokens to mitigate hallucination in large vision-language model. arXiv:2501.12206(2025)."},{"key":"44935_CR29","doi-asserted-by":"crossref","unstructured":"Jiang, Z. et al. Devils in middle layers of large vision-language models: Interpreting, detecting and mitigating object hallucinations via attention lens. arXiv:2411.16724 (2025).","DOI":"10.1109\/CVPR52734.2025.02328"},{"key":"44935_CR30","unstructured":"Lu, H. et al. Mitigating hallucination in multimodal reasoning via functional attention control. arXiv preprint arXiv:2510.10285 (2025)."},{"key":"44935_CR31","unstructured":"Zhang, X. et al. Seeing clearly by layer two: Enhancing attention heads to alleviate hallucination in lvlms. arXiv preprint arXiv:2411.09968 (2024)."},{"key":"44935_CR32","doi-asserted-by":"crossref","unstructured":"Xu, X. et al. Mitigating hallucinations in multi-modal large language models via image token attention-guided decoding. In Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers) 1571\u20131590 (2025).","DOI":"10.18653\/v1\/2025.naacl-long.75"},{"key":"44935_CR33","unstructured":"Kang, S., Kim, J., Kim, J. & Hwang, S.\u00a0J. See what you are told: Visual attention sink in large multimodal models. arXiv preprint arXiv:2503.03321 (2025)."},{"key":"44935_CR34","doi-asserted-by":"crossref","unstructured":"Liu, H., Li, C., Li, Y. & Lee, Y.\u00a0J. Improved baselines with visual instruction tuning. In Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition 26296\u201326306 (2024).","DOI":"10.1109\/CVPR52733.2024.02484"},{"key":"44935_CR35","unstructured":"Chiang, W.-L. et al. Vicuna: an open-source chatbot impressing gpt-4 with 90* chatgpt quality. 2, 6 https:\/\/lmsys.org (accessed 14 April 2023)."},{"key":"44935_CR36","unstructured":"Qwen et al. Qwen2.5 technical report (2025)."},{"key":"44935_CR37","doi-asserted-by":"crossref","unstructured":"Singh, A. et al. Towards VQA models that can read. In Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 8317\u20138326 (2019).","DOI":"10.1109\/CVPR.2019.00851"},{"key":"44935_CR38","doi-asserted-by":"crossref","unstructured":"Mathew, M., Karatzas, D. & Jawahar, C.\u00a0V. Docvqa: A dataset for VQA on document images. In Proceedings of the IEEE\/CVF Winter Conference on Applications of Computer Vision (WACV), 2200\u20132209 (2021). Accessed October 2025.","DOI":"10.1109\/WACV48630.2021.00225"},{"key":"44935_CR39","doi-asserted-by":"crossref","unstructured":"Schwenk, D., Khandelwal, A., Clark, C., Marino, K. & Mottaghi, R. A-okvqa: A benchmark for visual question answering using world knowledge. In  European Conference on Computer Vision (ECCV) 146\u2013162 (Springer, 2022). Accessed October 2025.","DOI":"10.1007\/978-3-031-20074-8_9"},{"key":"44935_CR40","doi-asserted-by":"crossref","unstructured":"Li, Y. et al. Evaluating object hallucination in large vision-language models. arXiv preprint arXiv:2305.10355 (2023).  Accessed October 2025.","DOI":"10.18653\/v1\/2023.emnlp-main.20"},{"key":"44935_CR41","doi-asserted-by":"crossref","unstructured":"Goyal, Y., Khot, T., Summers-Stay, D., Batra, D. & Parikh, D. Making the V in VQA matter: Elevating the role of image understanding in visual question answering. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 6904\u20136913 (2017). Accessed October 2025.","DOI":"10.1109\/CVPR.2017.670"},{"key":"44935_CR42","unstructured":"AI, X. Grok-1.5 vision preview (2024)."}],"container-title":["Scientific Reports"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.nature.com\/articles\/s41598-026-44935-1","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41598-026-44935-1.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41598-026-44935-1.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,7]],"date-time":"2026-05-07T11:02:00Z","timestamp":1778151720000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.nature.com\/articles\/s41598-026-44935-1"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,3,23]]},"references-count":42,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2026,12]]}},"alternative-id":["44935"],"URL":"https:\/\/doi.org\/10.1038\/s41598-026-44935-1","relation":{"has-preprint":[{"id-type":"doi","id":"10.21203\/rs.3.rs-7956158\/v1","asserted-by":"object"}]},"ISSN":["2045-2322"],"issn-type":[{"value":"2045-2322","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,3,23]]},"assertion":[{"value":"27 October 2025","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"16 March 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"23 March 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"The authors declare no competing interests.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Competing interests"}}],"article-number":"14576"}}