한인수 연구실

We focus on developing the fundamental algorithms and theoretical frameworks for fast, memory-efficient machine learning algorithms. Novel quantization techniques and pruning methods that significantly reduce model size and inference cost with minimal accuracy degradation. Hardware-aware inference algorithms and memory-efficient runtime systems for low-latency, high-throughput deployment on real devices. Attention approximation, KV cache optimization, and speculative decoding pushing the efficiency frontier of LLMs for practical deployment. We develop and analyze fast and memory-efficient machine learning algorithms — inspired by theoretical background— for efficient deployment of Artificial Intelligence models.