CIM-Pruner: A Dual-Mode Compute-In-Memory Macro for Efficient VLMs with Intra-Chunk Token Pruning and Merging

Zhuojun Han, Siqi He, Chixiao Chen, Haozhe Zhu. CIM-Pruner: A Dual-Mode Compute-In-Memory Macro for Efficient VLMs with Intra-Chunk Token Pruning and Merging. In IEEE International Symposium on Circuits and Systems, ISCAS 2026, Shanghai, China, May 24-28, 2026. pages 4681-4685, IEEE, 2026. [doi]

Abstract

Abstract is missing.