@article{indexcacheacceleratingsparseattentionvia, title = {IndexCache: Accelerating Sparse Attention via Cross-Layer Index Reuse}, author = {Yushi Bai and Qian Dong and Ting Jiang and Xin Lv and Zhengxiao Du and Aohan Zeng and Jie Tang and Juanzi Li}, year = {2026}, eprint = {2603.12201}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2603.12201}, }