宋社政，特任副教授，博士毕业于国防科技大学计算机学院。

研究方向：多模态大模型、多模态信息处理、多模态Agent、大模型持续学习。

博士毕业时获腾讯青云、字节筋斗云、深信服X-Star等多家人才计划，具备较强的业界能力和经验。

课题组现诚邀各位推免生及考研学生提前联系betterszsong@gmail.com。更多详细内容请见https://season1blue.github.io/

论文成果：

[1] Shezheng Song, et. al. Where Does Vision Meet Language? Understanding and Refining Visual Fusion in MLLMs via Contrastive Attention, CVPR, 2026. (CCF-A)

[2] Shezheng Song, et. al. How to bridge the gap between modalities: A comprehensive survey on multimodal large language model. TKDE, 2025.（CCF A, IF:10.4）

[3] Shezheng Song, et. al. A dual-way enhanced framework from text matching point of view for multimodal entity linking. AAAI, 2024. (CCF A)

[4] Shezheng Song, et. al. Leveraging Image as Compressed Visual Prompt and Hierarchical Visual Knowledge for Effective Image Utilization in MLLMs. AAAI, 2026. (CCF A)

[5] Ye Xiao, Shezheng Song*, et al. Beyond Token Stacking: Multi-scale Visual Modulation and Adaptive Fused-feature Calibration for MLLMs. ACM MM, 2026. (CCF A)
[6] Shezheng Song, Hao Xu, et al. HLoRA: Hierarchical Regularization on Combined Parameter Space for LoRA Fine-Tuning. EMNLP, 2026. (CCF B)
[7]Hao Xu, Shezheng Song*, et al. Small Models, Fewer Heads, More Forgetting: Sparse Attention Fine-Tuning for SLM Continual Learning. EMNLP, 2026. (CCF B)

[8] Shezheng Song, et. al. DIM: Dynamic Integration of Multimodal Entity Linking with Large Language Model. (PRCV), 2024. (CCF C)

[9] Shezheng Song, et. al. MOSA: A Large-scale Dataset for Multi-object Multimodal Aspect-based Sentiment Analysis. (TIP), 2026, (CCF A，Major Revision)

[10] Xiaopeng Li, Shasha Li, Shezheng Song, Jing Yang, Jun Ma, Jie Yu. PMET: precise model editing in a transformer. (AAAI), 2024. (CCF A)

[11] Xiaopeng Li, Shasha Li, Shezheng Song, Bin Ji, Xi Wang, Jun Ma, Jie Yu. Swea: Updating factual knowledge in large language models via subject word embedding altering. (AAAI), 2025. (CCF A)

[12] Xiaopeng Li, Shangwen Wang, Shasha Li, Shezheng Song, Bin Ji, Ma Jun, Jie Yu. Rethinking Residual Distribution in Locate-then-Edit Model Editing. (NeurIPS), 2025. (CCF A)

[13] Xiaopeng Li, Shasha Li, Shangwen Wang, Shezheng Song, Bin Ji, Huijun Liu, Jun Ma, Jie Yu, Identifying knowledge editing types in large language models. (SIGKDD), 2025. (CCF A)

[14] Tianwei Yan, Shan Zhao, Wentao Ma, Shezheng Song, Xinwang Liu, FRCL-MNER: A Finer Grained Rank-Based Contrastive Learning Framework for Multimodal NER. (TNNLS), 2025. (CCF B, IF=10.4)

[15] Chengyu Wang, Shan Zhao, Tianwei Yan, Shezheng Song, Wentao Ma, Kuien Liu, Meng Wang, Hierarchical label-enhanced contrastive learning for Chinese NER. (TNNLS), 2025. (CCF B, IF=10.4)