New KV cache compaction technique cuts LLM memory 50x without accuracy loss

MIT researchers developed Attention Matching, a KV cache compaction technique that compresses LLM memory by 50x in seconds — ...
Science Daily

MIT scientists strip cancer of its sugar shield

Scientists at MIT and Stanford have unveiled a promising new way to help the immune system recognize and attack cancer cells more effectively. Their strategy targets a hidden “off switch” that tumors ...
IEEE

A Structure Optimization Method Based on Stiffness Model for the Concentric Cable-Driven Manipulator With Three Segments

Abstract: The concentric cable-driven manipulator has characteristics of flexibility and slenderness when working in confined space. However, how to optimize a structure with appropriate stiffness ...
IEEE

Hardware Optimization Methods for High-speed Electrical Impedance Tomography System

Abstract: Electrical impedance tomography (EIT) is an advanced visualization modality. In recent years, high-speed devices are commonly employed in EIT systems to accomplish complex tasks and enhance ...